New insights into AtNBR1 as a selective autophagy cargo receptor in Arabidopsis

ABSTRACT

Selective autophagy, mediated by cargo receptors and recruiting specific targets to autophagosomes for degradation and recycling, plays an important role in quality control and cellular homeostasis in eukaryotes. The Arabidopsis AtNBR1 shares a similar domain organization with the mammalian autophagic receptors p62 and NBR1. We recently demonstrated that AtNBR1 functions as a selective autophagy receptor for the exocyst component AtExo70E2, a marker for the Exocyst-positive organelle (EXPO), which was achieved via a specific ATG8-AtNBR1-AtExo70E2 interaction in Arabidopsis. Here we further showed that nbr1 CRISPR mutants exhibit an early senescence phenotype under short-day growth conditions, which can be restored by complementation with expression of AtNBR1pro::AtNBR1-GFP in the mutant. Interestingly, in addition to the typical cytosolic and punctate patterns, YFP-AtNBR1 also exhibited a microtubule pattern particularly in the cortical layer. Treatments with the microtubule depolymerizer oryzalin but not the microfilament depolymerizer latrunculin B abolished the microtubule pattern and affected the vacuolar delivery of YFP-AtNBR1 upon autophagy induction. These results indicated that microtubules may be required for AtNBR1 to shuttle its cargos to the vacuole during plant autophagy. The present study thus sheds new light on the recognition and movement pattern of AtNBR1 in selective autophagy in Arabidopsis.

Selective autophagy is important to maintain intracellular homeostasis by specifically clearing and recycling of damaged organelles, invading pathogens and protein aggregates. The molecular machinery for selective autophagy requires autophagy receptors that recruit the targeted cargos to the growing autophagosomal membranes. Several cargo receptors for selective autophagy have been identified and characterized, including the sequestosome 1-like receptors p62/SQSTM1, NBR1, NDP52, NIX, OPTN, Sec62 and C53.^1–8 Among them, p62 and NBR1 are well-characterized receptors for autophagic degradation of mitochondria (mitophagy), peroxisomes (pexophagy), endoplasmic reticulum (reticulophagy), protein aggregates (aggrephagy) and pathogens (xenophagy).^{1,2,3,4,5,6,7,9} Structurally, similar to p62, NBR1 consists of a PB1 (Phox and Bem1p), a ZZ-type zinc finger, and a ubiquitin-associated (UBA) domain (Figure 1). The plant NBR1 homologs such as the Arabidopsis AtNBR1 and the tobacco NtJoka2 have a similar domain organization to those of mammals.^10,11 We have recently shown that the nbr1 CRISPR mutants exhibit typical autophagy-related early senescence phenotypes under short-day growth conditions,¹² suggesting its involvement in plant autophagy. Several studies on the plant NBR1 homologs have shown that NBR1 plays a conserved role in the plant response to abiotic stresses including sulfur deficit, heat, drought, salt and oxidative stress,^13–15 in which NBR1 mediated the degradation of ubiquitinated cargos or protein aggregates. In addition, NBR1 was shown to bind viral proteins and mediate their autophagy-dependent degradation during plant defense response.^16,17 Besides, AtNBR1 was also proposed to modulate abscisic acid (ABA) signaling via interacting with ABI3, ABI4 and ABI5.¹⁸ Interestingly, several other studies have indicated that the plant NBR1 might not be essential for the selective clearance of peroxisomes, mitochondria, or endoplasmic reticulum (ER) in plants.^14,19,20 However, the molecular mechanisms underlying plant NBR1 functions in selective autophagy remain elusive.

Figure 1.

Phylogenetic tree and conserved domain of NBR1 in plants, fungi and humans. Shown are the phylogenetic tree of the five NBR1/p62 from At, Nt, Hs, and Sm as well as their conserved domains. At, Arabidopsis thaliana (NP_194200); Nt, Nicotiana tabacum (NP_194200) Sm, Sordaria macrospora (XP_003346367); Hs, Homo sapiens (AAH01874.1, Q14596). PB1, Phox and Bem1p domain; ZZ, ZZ-type zinc finger domain; NBR1, neighbor of Brca1 gene; CC, coiled coiled domain; UBA, ubiquitin-associated domain

The most prevalent selective autophagy recognizing signal is the ubiquitination of cargos. The autophagic adapters p62 and NBR1 in mammalian cells interact with the ubiquitinated substrates to deliver them to autophagosomes for degradation. Although AtNBR1 has hybrid properties of both p62 and NBR1, our recent study showed that the binding of AtNBR1 to its cargo AtExo70E2, a marker for the Exocyst-positive organelle (EXPO), which was delivered to vacuole for turnover upon autophagy induction, is independent of the UBA domain,^12,21,22 which mediates the interaction with mono and polyubiquitin, suggesting a unique mechanism of AtNBR1-cargo interaction in Arabidopsis, which is distinct from the known mechanisms in mammalian cells.¹² Similarly, previous studies showed that the UBA domain of AtNBR1 did not bind to pathogen effectors including CaMV P4 and RPS6.^16,17 Thus, AtNBR1 might have different cargo recognition mechanisms in selective plant autophagy.

Several studies using the T-DNA insertional nbr1 mutants resulted in different plant phenotype observations.^13,23 As a first step to study the possible functions of AtNBR1 in selective autophagy, we have recently generated two new nbr1 CRISPR mutants (nbr1-c1 and nbr1-c2) for subsequent characterization of AtNBR1 function as a selective autophagy receptor for AtExo70E2.¹² Sanger sequencing showed that nbr1-c2 mutant has a premature stop codon at the 151aa (amino acid) site.¹² Unexpectedly, under short-day growth conditions (8-h/16-h, light/dark), the nbr1 CRISPR mutants exhibit an early senescence phenotype, although weaker than the autophagic mutant atg7-2 (Figure 2(a)). Such an early senescence phenotype in nbr1-c2 could be restored by complementation of AtNBR1pro::AtNBR1-GFP expression (Figure 2(a)). Since the truncated PB1 domain still exists in the nbr1 CRISPR mutants,¹² this new complementation study further confirmed that the accelerated early senescence phenotypes of the nbr1-c2 mutants under short-day growth conditions are indeed caused by AtNBR1 loss-of-function (Figure 2(a)). Interestingly, when the nbr1 T-DNA insertional mutant nbr1-2 was included in the phenotypic analysis, no obvious early senescence phenotype was observed in the nbr1-2 under short-day growth conditions.¹² To further investigate the endogenous protein levels of AtNBR1 in different nbr1 mutants, new antibodies targeting the N-terminal sequence from 200aa to 219aa (KKDISERTQTGRKPVNLNE) of AtNBR1 were generated and used in western blot analysis, in which a truncated 37-kDa AtNBR1 protein band was detected in the nbr1-2 mutant, but absent in the Col-0, nbr1-c1 and nbr1-c2 mutants.¹² Further whole genomic sequencing analysis covering all the T-DNA insertional sites could be performed to find out whether any transcripts contribute to the formation of the truncated AtNBR1 in the nbr1 T-DNA insertional mutants.

Figure 2.

Phenotypic analysis of nbr1 mutants under short day growth and nitrogen starvation conditions

(a). Leaf phenotypes of wild-type Col-0, nbr1-c2 and atg7-2. nbr1-c2 shows a hypersensitivity phenotype under short day growth conditions, which can be rescued by complementary expressing NBR1pro::NBR1-GFP, indicating the functional fusion of NBR1-GFP. Bar = 8 cm. (b). Phenotypic analysis from 7-day-old seedlings of Col-0, nbr1-c1, nbr1-c2 and atg7-2 under nitrogen starvation conditions, showing little difference among the WT and various mutants. Bar = 2 cm.(c). Quantification of chlorophyll contents of Col-0, nbr1-c1, nbr1-c2 and atg7-2 from (B). Error bars are the S.D. ns; not significant.

Interestingly, compared with the wild-type Col-0, nbr1-c1 and nbr1-c2 did not exhibit visible phenotypes under nitrogen starvation conditions (Figure 2(b,c)). In addition, there was no significant difference in chlorophyll contents among Col-0, nbr1-c1 and nbr1-c2. Since starvation-induced autophagy entails a bulk degradation of cytosolic materials,²⁴ these results indicate that AtNBR1 may not play an essential role in regulating starvation-induced autophagy (e.g. autophagy induction by nitrogen deprivation) in plants.

Recent confocal imaging analysis on root cells of transgenic Arabidopsis plant expressing YFP-AtNBR1 showed both cytosolic and punctate patterns of the fusion.¹² Interestingly, unexpectedly and intriguingly, when confocal imaging analysis was performed in the cortical layers of transgenic Arabidopsis root cells, a linear pattern of YFP-AtNBR1 was observed (Figure 3(a)). To find out if the linear pattern is related to microtubules, we also generated double transgenic lines co-expressing YFP-AtNBR1 and the microtubule marker mCherry-TUA5. As expected, YFP-AtNBR1 showed a partial colocalization with mCherry-TUA5 (Figure 3(b)). Furthermore, the microtubule localization of YFP-AtNBR1 was abolished after treatment with Oryzalin (a microtubule depolymerizer) (Figure 3(c)).

Figure 3.

YFP-AtNBR1 exhibits microtubule localization pattern in transgenic Arabidopsis

(a). Root cells of 5-day-old seedlings of transgenic Arabidopsis expressing YFP-AtNBR1 under normal conditions were subject to confocal imaging (at the cortical scanning layer), showing the microtubule localization pattern particularly in the cortical layer (indicated by arrows) of YFP-AtNBR1. Bar = 10 μm. (b). Root cells of 5-day-old seedlings of transgenic Arabidopsis co-expressing YFP-AtNBR1 and the microtubule marker mCherry-TUA5 under normal conditions were subject to confocal imaging (at the cortical scanning layer), showing the colocalization between YFP-AtNBR1 and mCherry-TUA5 (indicated by arrowheads as examples), along with the linear Pearson correlation coefficient (r_p = .41) and the nonlinear Spearman correlation coefficient (r_s = 0.32). Bar = 10 μm.(c). Root cells of 5-day-old seedlings of transgenic Arabidopsis co-expressing YFP-AtNBR1 and the microtubule marker mCherry-TUA5 were subjected to treatment by the microtubule depolymerizer oryzalin (10 μM), followed by confocal imaging (at the cortical scanning layer), showing that the oryzalin treatment disrupted the microtubule localization of YFP-AtNBR1, along with the linear Pearson correlation coefficient (r_p = .15) and the nonlinear Spearman correlation coefficient (r_s = 0.05). Bar = 10 μm.

The cytoskeleton plays a crucial role in the coordination of autophagosome formation in plants.^25,26 Disruption of microtubules restrains autophagosome formation upon autophagic induction.²⁷ However, how the cytoskeleton is coordinated for cargo trafficking and autophagosome formation remains poorly understood. In mammals, the evolutionarily conserved FW domain of NBR1 interacts with the microtubule-associated protein MAP1B to facilitate the movement of cargos destined to be degraded by autophagy.²⁸ In plants, NtJoka2, the homolog of AtNBR1 in tobacco, also colocalizes with both major cytoskeletal components: microtubules and microfilaments.²⁹ Interestingly, under autophagy induction conditions, the vacuolar delivery of YFP-AtNBR1 was specially abolished by treatment with Oryzalin, but not Latrunculin B (a microfilament depolymerizer) (Figure 4). Coincidentally, unlike mammalian cells, disruption of microfilaments does not affect basal and salt stress-induced autophagy in Arabidopsis.³⁰ These results indicate that microtubules may play roles in mediating vacuolar trafficking of AtNBR1 in selective autophagy in plants, further supporting the functional role of cytoskeleton in plant autophagy.^25–27 To provide genetic evidence in supporting this hypothesis, microtubules-related mutants could be adopted in future study to investigate whether microtubules rather than microfilaments play an essential role in mediating AtNBR1 vacuolar trafficking and function in plants.

Figure 4.

Microtubule integrity is essential for vacuolar delivery of YFP-AtNBR1 upon autophagy induction

(A) Root cells of 5-day-old transgenic Arabidopsis seedlings expressing YFP-AtNBR1 were subjected to (a) Mock (0.1% DMSO), (b) Latrunculin B (10 μM), and (c) Oryzalin (10 μM) for 8 h followed by confocal imaging. Arrows and arrowheads indicate examples of YFP-AtNBR1 in the cytoplasm. Bar = 20 μm. (B) Root cells of 5-day-old transgenic Arabidopsis seedlings expressing YFP-AtNBR1 were subjected to (a) Mock (0.1% DMSO), (b) Latrunculin B (10 μM), and (c) Oryzalin (10 μM) under autophagy induction conditions (+BTH at 100 μM +ConcA at 0.5 μM) for 8 h, followed by confocal imaging. Arrows and arrowheads indicate examples of YFP-AtNBR1 positive puncta in cytoplasm and vacuole, respectively. Bar = 20 μm. Conc A, Concanamycin A; BTH, Benzo-(1,2,3)-thiadiazole-7-carbothioic acid S-methyl ester.

In this short communication, we have demonstrated that AtNBR1-mediated selective autophagy plays an important role in the regulation of senescence in plants and that the vacuolar delivery of AtNBR1 during autophagy is dependent on the integrity of microtubules. In summary, Figure 5 shows the working model of the possible role and participation of microtubules in AtNBR1-mediated selective autophagy. Microtubules might be required for AtNBR1-mediated cargo trafficking, as well as long-distance transport of autophagosomes to the vacuole in plants. This model may service as another first step for future studies on the underlying mechanisms of selective autophagy in plants.

Figure 5.

Working model for the possible role and participation of microtubules in AtNBR1-mediated selective autophagy in plants

AtNBR1 has been shown to be selectively recruited various cargos like virus, protein aggregates and Exo70E2/EXPO to the autophagosomes for subsequent vacuolar delivery and vacuole degradation in plants.^8,9,14,16 The AtNBR1-mediated cargo transport toward autophagosomes might require the participation of microtubules via interaction with elusive microtubule-associated proteins. In addition, prior to autophagosome fusion with the vacuole, long-distance transport of autophagosomes might also rely on microtubules with unknown mechanisms.

Funding Statement

This work was partially supported by grants from the National Natural Science Foundation of China [31670179, and 91854201]; the Research Grants Council of Hong Kong [AoE/M-05/12, CUHK14130716, 14104716, 14102417, 14100818, 14101219, C4012-16E, C4033-19E, C4002-17G, G-CUHK410/19 and R4005-18] and CUHK Research Committee to L.J.