The ribosome-associated N-terminal acetyltransferase B coordinates global proteostasis and autophagy in plants by creating Ac/N-degrons

In a previous study, we reported delayed plant development in T-DNA insertion mutants of the NatB catalytic subunit gene NAA20 and the auxiliary subunit gene NAA25, both of which retained detectable levels of NatB complex transcripts and NAA25 protein⁹. To determine whether the complete loss of NatB in plants is lethal, as observed for NatB knockout mutants in animals¹⁴, we generated NAA20 null mutants using CRISPR/Cas9. Two independent mutations within the N-terminal coding region of NAA20 introduced frameshifts and premature stop codons, abolishing translation of the catalytic subunit and thereby fully inactivating NatB (Fig. 1a, Supplementary Fig. 1a). Both CRISPR/Cas9 knockout lines, designated naa20-cr1 and naa20-cr2, exhibited developmental retardation indistinguishable from the naa20-1 T-DNA mutant, including reduced rosette leaf size and fresh weight (30–40% growth delay) (Fig. 1b–d). These results indicate that complete loss of NAA20 is detrimental but not lethal in plants. Genetic complementation of naa20-cr1 with a NAA20–GFP fusion restored normal development (Fig. 1e, Supplementary Fig. 1b, c).

The N-acetylome was analyzed to evaluate the impact of NAA20 deletion on the NTA status of NatB substrates. N-terminomics identified 758 quantified non-redundant N-terminal peptides, of which 274 began at residue two due to iMet cleavage. Ninety-one N-termini retained the iMet, and 72 of these exhibited D/E/N/Q at position two, consistent with NatB substrate specificity (Supplementary Data). 1

In the wild type, 66 of 73 iMet-retaining proteins were fully N-acetylated. In contrast, both naa20-cr1 and naa20-1 mutants showed a pronounced reduction in NTA specifically for iMet-retaining peptides, while iMet-cleaved peptides (primarily acetylated by NatA) remained unaffected (Fig. 1f, g, Supplementary Fig. 1d, e). Notably, 33 iMet-retaining proteins displayed a more than 20% reduction in their Nt-acetylation in naa20 mutants compared with wild type (Supplementary Data 2). All of these proteins displayed the canonical NatB substrate sequence (ME, MD, MN or MQ). Twenty-five of the NatB substrates were found in both naa20 mutants; in those cases, the level of NTA was similarly decreased in both mutants.

Unexpectedly, not all NatB substrates in the mutants were unacetylated; many retained partial or full acetylation relative to wild type (Fig. 1f, g), suggesting the presence of a compensatory mechanism. Similar phenomena have been observed in vertebrates⁴⁰, potentially involving unidentified NATs or alternative known NATs acting on iMet-initiated peptides (e.g., NatC/E/F). To test the latter hypothesis, we overexpressed the catalytic subunits of NatC (NAA30), NatE (NAA50), and NatF (both long and short NAA60 isoforms). However, none of them restored NTA levels or rescued the developmental defects in the naa20-cr1 mutant (Fig. 1h, i, Supplementary Fig. 2).

Recently, the role of NatA in stabilizing the proteome has been revealed in various species¹⁷. NatB is responsible for the NTA of more approximately 20% of the plant proteome. To investigate its role in protein turnover and proteome stability, we measured the chymotrypsin-like protease activity of the proteasome in soluble protein extracts of the wild type and the natb mutants by monitoring the release of fluorescent AMC from the model substrate Z-Leu-Leu-Leu-AMC. We found that the proteasome activity was significantly reduced in the natb mutants compared with wild type (Fig. 2a). Consistent with this finding, both overall and K48-linked polyubiquitination levels showed reduced levels in the soluble protein fraction of natb mutants (naa20-cr1 and naa25-1) compared with wild type when proteasome activity was inhibited by MG132 (Fig. 2b, c, Supplementary Fig. 3a, b). These findings imply a lower in vivo degradation rate by the UPS in natb mutants. Quantification of protein translation rates by feeding of isotope-labeled amino acids or azidohomoalanine showed that translation rates were also reduced in the natb mutants (Fig. 2d, Supplementary Fig. 3c, d).

To determine whether the lowered protein turnover in naa20-cr1 specifically leads to the accumulation of NatB substrates, we performed a quantitative proteomic analysis. A total of 3788 proteins were identified, with NatA and NatB targeting 61% and 19% of them, respectively (Fig. 2e, Supplementary Data 3). In the wild type, only 16% of stabilized proteins are NatB substrates, indicating that NatB substrates are not enriched in the stabilized protein fraction (as 19% of all identified proteins are NatB substrates). In the natb mutant, 30% of stabilized proteins are NatB substrates, representing a substantial enrichment of NatB substrates in this fraction when NatB is absent. The percentage of stable NatB proteins increased from 16% in the wild type to 30% in the natb mutant (Fig. 2e). This finding demonstrates that NAA20 is critical for controlling the stability of its substrate in plants. A gene ontology (GO) analysis using agriGO v2.0⁴¹ revealed that the proteins with increased abundance in the naa20-cr1 mutant were mainly enriched in translation and amino acid metabolic process pathways (Fig. 2f), which may be explained by an unsuccessful attempt to compensate for the impaired translation rate in naa20-cr1 (Fig. 2d). Proteins with reduced abundance in the naa20-cr1 mutant were mainly enriched in photosynthesis and protein synthesis pathways (Fig. 2f), reinforcing the proposed role of NAA20 in determining plant protein homeostasis and plant growth.

To further investigate the mechanism by which NatB regulates plant proteome stability, we performed an ubiquitin capture (Ubi capture) approach to enrich ubiquitinated proteins. The Ubi capture pulldown was followed by mass spectrometry to identify and classify differentially polyubiquitinated proteins in the wild type and naa20-cr1. The Ubi capture results confirmed dramatically reduced polyubiquitination levels in naa20-cr1 compared with wild type (Fig. 3a). Consistent with this, the vast majority (310 out of 314) of significantly changed polyubiquitinated proteins identified by mass spectrometry were more abundant in the wild type than in naa20-cr1 (Supplementary Data 4), suggesting that NatB destabilizes the plant proteome. Classification of NAT substrates for less ubiquitinated proteins in naa20-cr1 showed that NatB substrates were not significantly enriched, with only 16% of the proteins identified as NatB substrates, while NatA substrates represented 68% of the less ubiquitinated proteins in naa20-cr1 (Fig. 3b). GO analysis using agriGO v2.0⁴¹ revealed that a large number of less polyubiquitinated proteins in the naa20-cr1 mutants are involved in protein hydrolysis, translation, and response to abiotic stimuli (Fig. 3c). Among these, 86% of the identified proteins involved in the translation process are NatA substrates, whereas only 4% of them are NatB substrates (Fig. 3d). However, 42% of the identified proteins involved in the ubiquitin-dependent protein catabolic process are NatB substrates (Fig. 3e). These findings suggest that NatB may not directly regulate the plant proteome by modulating the stability of all NatB substrates. Alternatively, NatB-mediated substrates may play a role upstream of the protein homeostasis cascade.

Next, we examined the protein abundance of six NatB substrates (V-ATPase subunit, RPN12a, MAT3, PBA1, HSP101, and NBR1) in the wild type and naa20-cr1 by immunological detection. Among these, V-ATPase subunit and RPN12a have been experimentally confirmed to have reduced NTA levels in the naa20 mutants (Supplementary Data 2), whereas the other four candidates are predicted NatB substrates based on their sequences. The results suggested that only NBR1, a selective autophagy cargo receptor, showed higher protein abundance in naa20-cr1 compared with wild type (Fig. 3f), implying that NatB may not directly affect the protein stability of all individual NatB substrates.

NBR1 was not quantified by the proteomics approach (Fig. 2e). To further investigate how NatB regulates the protein abundance of NBR1, we excluded the possibility of elevated NBR1 transcripts in mutants by RT-qPCR (Supplementary Fig. 4a). In vitro acetylation assays demonstrated that the endogenous NatB complex, immunoprecipitated from a NAA20-GFP/naa20-cr1 line using GFP beads, catalyzes the acetylation of the NBR1 N-terminus. The N-terminally acetylated NBR1 (Ac-NBR1) and the mutagenized NBR1^E2P N-terminus served as negative controls and were not accepted as substrates by NatB (Supplementary Fig. 4b, c). Next, we applied cycloheximide (CHX) chase experiments and found that NBR1 was significantly slower degraded in naa20-cr1 than in the wild type, implying that NatB-mediated NTA promotes degradation of soluble NBR1 (Fig. 3g). To further support this hypothesis, we examined the degradation rate of the NBR1^E2P variant, which cannot be acetylated in vivo, and found that the degradation rate of NBR1^E2P was indeed significantly reduced compared with the wild-type NBR1 (Fig. 3h). Previous studies have shown that NBR1 is a selective cargo receptor for autophagy and is degraded by autophagy in the vacuole. However, we found that treatment with MG132, an inhibitor of the proteasome, also caused its protein accumulation, and after combining Concanamycin A (ConA) to block autophagy, the amount of soluble NBR1 accumulated further (Supplementary Fig. 4d). These findings imply that the UPS regulates the amount of soluble NBR1 in an NTA-dependent manner, making it available to serve as a membrane-associated autophagy cargo receptor. To further support this conclusion, we analyzed NBR1 degradation dynamics in naa20-cr1, when autophagy, the UPS or both are suppressed by MG132 and/or ConA treatments. When protein synthesis was inhibited by Cycloheximide (CHX), both MG132 and ConA alone inhibited NBR1 degradation, with MG132 exhibiting a more pronounced effect in plants grown under optimal growth conditions. When MG132 and ConA were applied together, NBR1 degradation was completely inhibited, indicating that NBR1 is degraded via both the UPS and autophagy pathways in vivo (Supplementary Fig. 4e). In line with this hypothesis, we found that NBR1 was also degraded in vitro in a membrane-free system. The degradation rate of NBR1 in this membrane-free system was slower in the naa20-cr1 and could be inhibited in the wild type by the addition of MG132 (Supplementary Fig. 4f). Taken together, these results demonstrate that, besides autophagic degradation, NBR1 undergoes UPS-mediated degradation by creating Ac/N-degron.

N-terminal acetylation can influence not only protein stability but also subcellular localization and protein–protein interactions. Next, we investigated the localization of NBR1 using stable NBR1-mCherry-GFP/WT and NBR1-mCherry-GFP/naa20-cr1 transgenic lines. Microscopy results revealed that in wild-type root tip cells, NBR1 shows a cytosolic and a punctate localization pattern, which is consistent with previous studies⁴². In naa20-cr1 mutant, the localization pattern of NBR1 remains unchanged, indicating that the absence of NTA does not affect NBR1 localization (Supplementary Fig. 4g). To investigate whether loss of N-terminal acetylation alter the interaction between NBR1 and ATG8, we performed Co-IP assay using stable NBR1-HA/WT and NBR1-HA/naa20-cr1 transgenic plants. The results showed that NBR1 still interacts with ATG8 in naa20-cr1, suggesting that the absence of NTA does not affect the interaction between NBR1 and ATG8 (Supplementary Fig. 4h).

Previous reports have demonstrated that a lowered proteasome abundance could be caused by enhanced autophagy⁴³. To investigate whether the lower protein turnover via the UPS in the NatB mutant might be partially a consequence of a higher autophagy level in NatB-deficient mutants, we examined the K63-type polyubiquitination signals in the wild type and naa20-cr1 plants after inhibition of autophagy by ConA. In contrast to the K48-linked polyubiquitination (Fig. 2c), the K63-linked polyubiquitination was significantly increased in naa20-cr1 compared with wild type, indicating a higher autophagic flux in the naa20-cr1 mutant (Fig. 4a, b). This conclusion was further supported by the detection of autophagy-related proteins ATG8 and ATG3, which are essential for autophagosome formation, in natb mutants and the wild type. The abundance of these ATG proteins was significantly increased in natb mutants (Fig. 4c–e). This increase also applied to ATG8-PE, a post-translationally lipidated proteoform of ATG8 covalently conjugated to phosphatidylethanolamine, which is essential for autophagosome membrane expansion (Fig. 4c), and the ATG5-ATG12 conjugated complex, which is essential for autophagosome formation by acting as an E3-like enzyme that promotes ATG8 lipidation (Fig. 4e). Neither ATG8–PE nor ATG5-ATG12 conjugated complex was detected in the atg5-1 mutant (Fig. 4c, d). Real-time PCR analysis revealed no significant increase in the expression levels of these genes in the mutant, suggesting that the increase of ATG proteins may be associated with changes in post-translational modifications (Fig. 4f).

Next, we applied the autophagy marker GFP-ATG8a to analyze the autophagic flux in the wild type and the naa20-cr1 mutant. Non-invasive confocal laser scanning microscopy of GFP-ATG8a/naa20-cr1 and GFP-ATG8a/wild type demonstrated that after selective inhibition of autophagic flux by ConA application, darkness-treated naa20-cr1 mutants accumulated significantly more autophagic bodies in their vacuole than the wild type (Fig. 4g, h). Moreover, to further substantiate the increased autophagic flux in the naa20-cr1 mutant, a GFP-cleavage assay was performed. Darkness treatment caused a substantially stronger induction of GFP release from GFP-ATG8a in the naa20-cr1 when compared with the wild type as indicated by the higher GFP to GFP-ATG8a ratio in naa20-cr1 at all times of darkness treatment (4, 8, and 12 h, Fig. 4i). Taken together, these data demonstrate a higher autophagic flux in the absence of NatB.

The connection between N-terminal protein acetylation and autophagy remains unaddressed in plants so far. To verify whether the constitutively elevated autophagic flux could cause more efficient physiological adaptation to stress conditions in naa20 mutants, we subjected the wild type, two naa20 mutants, and the autophagy-deficient mutant atg5-1 to elongated darkness stress. Both naa20 mutants exhibited significantly higher survival rates compared with the wild type upon darkness treatment. As expected, the autophagy mutant atg5-1 was susceptible to prolonged darkness stress (Fig. 5a, b). Consistent with this, the natb mutants exhibit delayed senescence after growth in short-day condition for 10 weeks and transfer to long-day condition for 2 weeks (Supplementary Fig. 5a). Furthermore, the naa20-cr1 showed higher resistance and more fresh weight accumulation upon nitrogen and sulfur starvation conditions. The autophagy-deficient mutant performed worse than the wild type under both nutrient deficiencies. However, under control conditions, naa20-cr1 and atg5-1 displayed similarly decreased growth compared to the wild type (Fig. 5c, d). These results suggest that, unlike in yeast, where NatB is critical for autophagy induction³¹, NatB functions as a negative regulator of autophagy in plants.

To further investigate the role of the induced autophagic flux in natb mutants, we crossed naa20-cr1 with two atg mutants, atg5-1 and atg7-1, respectively. We found that blocking autophagy did not rescue the growth defects of naa20-cr1 under optimal growth conditions (Supplementary Fig. 5b, c). Nevertheless, both naa20-cr1 atg5-1 and naa20-cr1 atg7-1 double mutants lost the resistance to prolonged darkness (Fig. 5e, f). Inactivation of autophagy in the naa20-cr1 atg double mutants restored wild-type levels of polyubiquitination, which is in agreement with the hypothesis that induction of autophagy is partially responsible for the disturbed protein turnover in naa20-cr1 (Fig. 5g, h). In support of a substantial crosstalk between autophagy and the UPS, the autophagy-deficient single mutants exhibited higher proteasome activity and polyubiquitination levels compared with the wild type (Fig. 5g, h, Supplementary Fig. 5d), demonstrating that the UPS is generally activated after the blockage of autophagy. To verify that crossing of atg5-1 into naa20-cr1 indeed inhibited autophagy in the double mutant, we crossed the GFP-ATG8a marker into this line and demonstrated the absence of autophagosome formation in this mutant (Supplementary Figs. 5e, f). Taken together, these results demonstrate that NAA20 regulates resistance to carbon starvation through autophagy and strongly suggest that constitutively induced autophagy may contribute to the downregulation of UPS in NatB-deficient plants.

To understand how the darkness stress-resistant phenotype is induced in natb mutants, we crossed naa20-cr1 and nbr1-2 and found that the naa20-cr1 nbr1-2 double mutant is still resistant to darkness (Supplementary Fig. 6a, b), implying that the accumulation of NBR1 in naa20-cr1 does not contribute to the darkness resistance phenotype of naa20-cr1. We then examined the N-terminal sequences of autophagy-related proteins, as well as those related to the upstream regulatory network of autophagy induction. We found that the two catalytic subunits of the plant energy sensor SnRK1, KIN10 and KIN11, are predicted to be NatB substrates. Thus, we detected the activation status of SnRK1 and found that pKIN10 and pKIN11, two proteoforms that are more phosphorylated at the kinase activation loop, accumulated in naa20-cr1, demonstrating that SnRK1 activity is substantially enhanced in the naa20-cr1 mutant (Fig. 6a, Supplementary Fig. 6c). Consistent with the activation of SnRK1, we found that the transcriptional levels of energy-related genes downstream of SnRK1 (DIN1, DIN6 and DIN10) were induced in natb mutants. Remarkably, the transcription factor bZIP63, which is directly phosphorylated by SnRK1 and controls transcription of these DIN genes, is not induced, suggesting that DIN gene activation is caused by post-translational modification of bZIP63. At the same time, endogenous activity of the sensor kinase TARGET OF RAPAMYCIN (TOR), as determined by the ratio of phosphorylated-RPS6 to RPS6, was suppressed in natb mutants, which may explain the decreased translation and lowered growth of these mutants (Supplementary Fig. 6d, e). Based on these findings, we hypothesized that the absence of NTA on KIN10 and KIN11 directly causes the accumulation of pKIN10 and pKIN11, thereby enhancing SnRK1 activity in natb mutants. In vitro acetylation experiments provided evidence that NatB indeed acetylates the major proteoforms of KIN10 and KIN11 (Fig. 6b). In contrast, the low-abundant proteoform of KIN10 (KIN10L), which starts with an MF, is not acetylated by NatB in vitro. Destruction of the canonical NatB substrate recognition signal in the mutagenized major protoeforms of KIN10^D2P or KIN11^D2P impaired acetylation of the major SnRK1 isoforms by NatB (Fig. 6b). Since we found that several NatB substrates accumulated in natB mutants (Fig. 2e), we tested the abundance of KIN10 and KIN11 in natb and the wild type by immunological detection. Only KIN11 but not KIN10 significantly accumulated in naa20-cr1 (Fig. 6c, d). In this context, it is remarkable that KIN10 and KIN11 exhibit 89.2% similarity, and that only three protein domains show greater variability between the proteins (Supplementary Fig. 6f). The domain with the highest difference between both proteins is the N-terminus, suggesting that the distinct N-terminal sequences of KIN11 (MDHSSNRFGN¹⁰-) and KIN10 (MDGSGTGSRS¹⁰-) contribute substantially to the observed differences in accumulation.

To provide direct evidence for the specific stabilization of the non-acetylated KIN11 protein in natb mutants, we firstly determined the in vivo NTA levels of KIN11-HA in wild-type and naa20-cr1 background by immunoprecipitation followed by mass spectrometry and found that the loss of NatB resulted in a two-fold lower KIN11 acetylation in naa20-cr1 when compared to wild type (Supplementary Fig. 6g, h). We then excluded the possibility of elevated KIN11 transcripts in the natB mutants as the reason for KIN11 accumulation (Supplementary Fig. 6i). CHX time-chase experiment revealed that the degradation of KIN11 could be inhibited by the proteasome inhibitor MG132 (Supplementary Fig. 6j) and that the degradation rate of KIN11 was significantly lower in the naa20-cr1 mutant than in the wild type (Fig. 6e), strongly suggesting that NTA of KIN11 leads to its degradation in a UPS-dependent manner. To further support this hypothesis, we examined the degradation rate of the KIN11^D2P variant, which cannot be acetylated in vivo, and found that the degradation rate of KIN11^D2P-GFP was significantly decreased when compared with the regular KIN11-GFP fusion protein starting with MD (Fig. 6f). Moreover, immunoprecipitation of KIN11-GFP followed by immunological detection of polyubiquitination revealed substantially higher accumulation of polyubiquitylated KIN11-GFP in the wild-type than in the naa20-cr1 (Fig. 6g). Furthermore, the non-acetylated KIN11^D2P-GFP was less polyubiquitinated than the acetylated KIN11-GFP in the wild type, demonstrating that the NTA status of the KIN11-GFP determines its degradation rate by the UPS (Fig. 6g). In summary, our results uncover that decreased NTA of KIN11 impairs its degradation by the UPS, which leads to the accumulation of the activated KIN11 protein. In agreement with our hypothesis that KIN10 is not destabilized in natb mutants due to its significantly different N-terminus when compared to KIN11, we found that KIN10 is not destabilized in naa20-cr1 (Supplementary Fig. 6k).

Previous studies have shown that despite the substantial redundancy between KIN10 and KIN11, both isoforms also have specialized functions^44,45. Motivated by these reports, we investigated the darkness resistance of kin10 and kin11 mutants and found that only kin11, but not kin10, was more sensitive to darkness treatment (Supplementary Fig. 7a, b).

To investigate whether the elevated KIN11 protein level is responsible for the increased darkness resistance of naa20-cr1, we crossed naa20-cr1 with kin10 and kin11. Remarkably, the darkness resistance of the naa20-cr1 kin11 double mutant was attenuated compared with that of naa20-cr1. In contrast, inactivation of KIN10 in the naa20cr-1 kin10 double mutant did not affect the darkness resistance of the NatB-deficient mutant (Fig. 7a, b). Correspondingly, two independent KIN11-overexpressing lines exhibited higher survival rates than the wild type upon darkness treatment, suggesting a role for KIN11 in regulating carbon starvation response in plants (Supplementary Fig. 7c, d). Detection of the phosphorylated SnRK1 catalytic subunits showed that the phosphorylated KIN11 was absent in the naa20-cr1 kin11 double mutant, while the kin11 and naa20-cr1 kin11 mutants still showed comparable levels of phosphorylated KIN10 that were higher than those of the wild type. These findings demonstrate that the specific activation of KIN11 is critical for the darkness resistance of naa20-cr1 (Fig. 7c). However, the knockout of KIN11 fails to suppress the elevated ATG proteins in naa20-cr1 kin11 double mutant, implying that activation of either KIN10 or KIN11 in the natb mutants is sufficient for autophagy induction under non-stressed conditions (Supplementary Fig. 7e, f). Taken together, we found that NatB regulates the interplay between UPS and autophagy in plants by fine-tuning the SnRK1 activity as well as the abundance of autophagy-associated proteins, which in turn coordinates plant growth and development and nutrient stress accordingly.

In summary, we found that the absence of NTA on many NatB substrates causes their lowered degradation by the UPS (Figs. 2c, 3g, 3h, 6e–g), which causes them to accumulate (Figs. 2e, 3f, 6c). A decreased global protein turnover and lowered UPS accompany this (Figs. 2a–d, 3a). However, we could not solely explain the decreased polyubiquitination levels in natb mutants by the lowered polyubiquitination of NatB substrates (Fig. 3b). Instead, we found that decreased NTA of the catalytic subunits of the energy sensor SnRK1 (KIN10 and KIN11) caused their activation (Fig. 6a), which resulted in a constitutive induction of autophagy in NatB-deficient mutants (Figs. 4, 7a), most likely contributing to compensatory general down-regulation of the UPS (Figs. 2a-c, 3a, 3b). Finally, we uncovered the molecular mechanism of the darkness stress resistance of NatB-deficient mutants, which is caused by the specific accumulation of activated KIN11 protein (Figs. 5a, 5b, 5e, 5f, 7a, 7b). Our findings demonstrate that NatB controls the stability of diverse NatB substrates by the UPS and attenuates autophagy levels in plants by regulating KIN11 abundance via the generation of an Ac/N-degron (Figs. 3g, 3h, 6e–g, Supplementary Fig. 4). Consequently, NatB is a critical regulator of the UPS-autophagy interplay, which explains its resistance to diverse nutrient limitation stresses (Fig. 7d).

Arabidopsis thaliana ecotype Col-0 was used as the wild type. The T-DNA insertion mutants naa20-1 (SALK_027687) and naa25-1 (GK-819A05) were previously published in ref. ⁹. The T-DNA insertion mutants kin11 (WiscDsLox320B03) and kin10 (GABI_579E09) were previously published in ref. ⁶⁰. The T-DNA insertion mutants nbr1-2 (GABI_246 H08), agt5-1 (SAIL_129_B07), atg7-1 (N39995↗) were obtained from the Nottingham Arabidopsis Stock Centre (www.arabidopsis.info↗). The pUb:GFP-ATG8a lines in both wild type and agt5-1 backgrounds were previously published in ref. ⁶¹. The pUb:GFP-ATG8a lines in naa20-cr1 and naa20-cr1 agt5-1 backgrounds were generated by genetic crossing naa20-cr1 with the respective pUb:GFP-ATG8a lines. The naa20-cr1 agt5-1, naa20-cr1 agt7-1, naa20-cr1 nbr1-2, naa20-cr1 kin11 and naa20-cr1 kin10 double mutants were generated by genetic crossing the naa20-cr1 with respective single mutants.

Seeds were sown in the pot filled with well-watered turf mixture (Ökohum, Herbertingen) supplemented with 10% [v/v] vermiculite and 2% [v/v] quartz sand. Then the seeds were vernalized in the dark at 4 °C for 2–3 days and transferred to short-day conditions (8.5 h light per day, light intensity: 100 μmol m⁻² s⁻¹; day temperature: 22 °C; night temperature: 18 °C; relative humidity: 50%) in plant growth chambers (ThermoTec). For seeds production or crossing, eight-week-old plants were transferred to long-day growth chambers (16-hour light per day; other conditions were the same as in short-day conditions).

To generate NAA20 CRISPR/Cas9 construct, single-guide RNA (sgRNA) targeting the first exon of NAA20 was designed (Supplementary Data 5) and cloned according to the protocol⁶². Selection of NAA20 CRISPR/Cas9 seedlings was performed by applying Hygromycin B (25 mg/L) on solid ½ MS medium plates. The surviving plants were sequenced using specific primers to identify mutations (Supplementary Data 5). The CRISPR/Cas9 construct was eliminated from naa20-cr mutants by self-pollination and segregation.

To generate the NAA20-GFP, NBR1-HA, NBR1-mCherry-GFP, KIN11-GFP, KIN11^D2P-GFP, KIN11-HA tagged constructs and NBR1^E2P, NAA30, NAA50, NAA60S, NAA60L overexpression constructs, the original and variant cDNA sequences of the respective genes were amplified using the PCRBIO HiFi Polymerase (PCRBIOSYSTEMS) with specific primers (Supplementary Data 5). The resulting PCR fragments were cloned into the GreenGate cloning system under the control of the Ubiquitin promoter. Final vectors were sequenced to confirm sequence identity (EUROFINs) and then stably transformed into respective plants by A. tumefaciens–mediated floral dipping. Selection of one-week-old soil-grown transgenic plants was performed by spraying glufosinate-ammonium solution (0.2 g/liter) three times within 6 days. The surviving plants were genotyped by PCR using specific primers or sequenced to confirm the variants and backgrounds (Supplementary Data 5).

Total RNA was isolated from 100 mg of leaf materials of soil-grown plants using the Universal RNA Kit (Roboclon). Subsequently, first-strand cDNA synthesis was carried out with the Fast Gene Scriptase II cDNA Synthesis Kit (NIPPON Genetics) following the manufacturer’s instructions to generate templates for both RT-qPCR and RT-PCR analyses. RT-qPCR was performed using the qPCRBIO SyGreen Mix Lo-ROX (PCR Biosystems) following the manufacturer’s instructions. RT-PCR was performed using the FastGene Taq 2× Ready Mix (NIPPON Genetics). The specific primer pairs used are listed in Supplementary Data 5. The expression levels were normalized to the expression of Actin7 (At5g09810) or PP2A (AT1G69960).

The soluble proteins were extracted from 100 mg of leaf materials from six-week-old soil-grown plants under short day condition with 300 μl of extraction buffer (50 mM HEPES pH 7.4, 10 mM KCl, 1 mM EDTA, 11 mM EGTA, and 10% [v/v] glycerol) supplemented with 10 mM DTT, 0.5 mM PMSF and 1 x complete protease inhibitor cocktail (Roche). For the detection of RPS6 and autophagy-related proteins, NuPage sample buffer (150 mM Tris-HCl pH 8.5, 2% LDS, 1 mM EDTA, 10% [v/v] glycerol, 0.22 mM Coomassie Blue G250, 0.166 mM Phenol Red and 50 mM DTT) was used to extract total protein. Subsequently, proteins were separated by SDS-PAGE or Tricine-SDS-PAGE (for ATG8-PE separation) and blotted onto PVDF membranes. Antibodies against NAA25 (1:5,000)⁹, NAA50 (1:5,000)⁵⁷, GFP (ChromoTek, Pabg1-100, 1:2,000 and Santa-Cruz Biotechnology, sc-9996 HRP, 1:2,000), K48-Ubi (Cell signaling, 8081, 1:5,000), K63-Ubi (Cell signaling, 5621, 1:1,000), Neutravidin-HRP (Invitrogen, 31001, 1:10,000), mono- and poly-ubiquitination (Enzo life Sciences, ENZ-ABS840HRP, 1:10,000), RPN12a (Agrisera, AS19 4268, 1:2,000), V-ATPase (Agrisera, AS07 213, 1:2,000), HSP101 (Agrisera, AS07 253, 1:2,000), SAM1-4 (Agrisera, AS16 3148, 1:2,000), PBA1 (Agrisera, AS19 4260, 1:2,000), ATG3 (Agrisera, AS19 4275, 1:3,000), ATG5 (Agrisera, AS15 3060, 1:2,000), ATG8 (Agrisera, AS14 2769, 1:2,000), RPS6 (Cell signalling, 2317, 1:2,000), p-RPS6 (1:5,000)⁶³, KIN10 (Agrisera, AS21 4581, 1:2,000), KIN11 (Agrisera, AS10 920, 1:1,000), Phospho-AMPKα (Thr172) Antibody (Cell signaling, 2531S, 1:2,000) or the secondary anti-rabbit IgG-HRP (Agrisera, AS09 602, 1:25,000) were diluted in 1× Tris-buffered saline (50 mM Tris, pH 7.6, 150 mM NaCl, and 0.05% [v/v] Tween 20) supplemented with 0.5% [w/v] BSA. Antibodies against NBR1 (Agrisera, AS14 2805, 1:5,000) was diluted in 1×TBST supplemented with 8% [w/v] nonfat milk. The resulting signals were quantified with the Fiji image processing application. All the uncropped blots are provided in the Source Data file.

Leaf materials from six-week-old soil-grown plants or seedlings from ½ MS plates under short day condition were used to determine the protein degradation rate. For in vitro cell-free protein degradation assay, the soluble proteins were extracted from leaf materials by adding two volumes of extraction buffer (25 mM Tris–HCl pH 7.5, 10 mM NaCl, 10 mM MgCl2, 4 mM PMSF, 5 mM dithiothreitol, and 10 mM ATP). Supernatants were cleared by centrifugation at 20,000 g for 20 min at 4 °C, and incubated at 22 °C for indicated time. For CHX time chase assay, protein synthesis was inhibited by floating leaf discs or seedlings from ½ MS plates on ½ Hoagland medium (2.5 mM Ca(NO₃)₂, 0.5 mM MgSO₄, 2.5 mM KNO₃, 0.5 mM KH₂PO₄, 4 μM Fe-EDTA, 25 μM H₃BO₃, 2.25 μM MnCl₂, 1.9 μM ZnCl₂, 0.15 μM CuCl₂, 50 nM (NH₄)₆Mo₇O₂₄, and 0.6% (w/v) microagar pH 5.8) supplemented with 100 μM CHX (Sigma Aldrich) for indicated time under the light at 22 °C while shaking at 60 rpm/min. MG132 (50 μM) or ConA (1 μM) were applied to inhibit UPS and autophagy-mediated degradation. The collected samples were subjected to SDS-PAGE and immunoblot analysis using specific antibodies.

HA magnetic beads and GFP agarose beads based immunoprecipitation was performed following the manufacturer’s instructions. Proteins were extracted from transgenic plants expressing the respective HA or GFP fusion proteins with extraction buffer (50 mM Tris–HCl pH 7.5, 150 mM NaCl, 10% glycerol, 1 mM PMSF and 1 × complete protease inhibitor cocktail and 0.1% Nonidet P-40). To confirm the presence of the NatB complex in the IP products of NAA20-GFP/naa20-cr1 plants, IP products were detected with GFP and NAA25 antibodies. To detect polyubiquitinated KIN11-GFP and KIN11^D2P-GFP in KIN11-GFP/WT, KIN11-GFP/naa20-cr1 and KIN11^D2P-GFP/WT plants, 50 μM of MG132 was applied to leaf discs and incubated for 5 h to accumulate polyubiquitination. The respective IP products were detected with GFP and ubiquitin antibodies. To investigate the interaction between NBR1 and ATG8, stable NBR1-HA/WT and NBR1-HA/naa20-cr1 transgenic plants were used.

SDS-PAGE gels containing HA-immunoprecipitated products from KIN11-HA/WT and KIN11-HA/naa20-cr1 were manually excised after coomassie staining. The in-gel Trypsin digestion was performed following by Nanoflow LC-MS2 analysis using the Vanquish Neo coupled to an Orbitrap Eclipse (Thermo Fisher Scientific). An in-house packed analytical column (75 µm x 200 mm, 1.9 µm ReprosilPur-AQ 120 C18 material was used. The fragment ion mass tolerance was set to 0.02 Da, and the parent ion mass tolerance to 10 ppm. The following variable modifications were allowed: Oxidation (O), Met-loss (M), Acetyl (N-terminus), Deamidation (N, Q), whereas Carbamidomethylation (C) was set as a fixed modification. Peptide quantification was done using a precursor ion quantifier node with the Summed Intensity method set for protein abundance calculation.

The in vitro NTA assay was performed as previously described in ref. ⁶⁴. The purified NatB complex using GFP agarose-beads from NAA20-GFP/naa20-cr1 plants was used. Samples without NatB complex or with heat-inactivated NatB complex were used for negative controls. The absorbance of negative controls without NatB complex was averaged and subtracted from the absorbance of the samples. The product concentration was calculated using the Lambert Beer equation and the extinction coefficient of TNB2-(13700/M cm). The following synthetic peptides (GenScript, >95% purity) were used: NBR1: [H]MESTANALVVRVSYGGVLRR[OH]; NBR1^E2P: [H]MPSTANALVVRVSYGGVLRR[OH]; Ac-NBR1: ac-MESTANALVVRVSYGGVLRR[OH]; KIN11: [H]MDHSSNRFGNNGVESILPNY[OH]; KIN11^D2P: [H]MPHSSNRFGNNGVESILPNY[OH]; ac-KIN11: ac-MDHSSNRFGNNGVESILPNY[OH]; KIN10S: [H]MDGSGTGSRSGVESILPNY[OH]; KIN10^D2P: [H]MPGSGTGSRSGVESILPNY[OH]; ac-KIN10S: ac-MDGSGTGSRSGVESILPNY[OH] and KIN10L: [H]MFRRVDEFNLVSSTIDHRI[OH]. The lysine residue in NBR1 peptides was replaced by arginine residues to minimize any potential interference by Nϵ-terminal acetylation.

Free N-termini groups in foliar proteins from six-week-old soil-grown plants were labeled with 0.5 mM NBD-Cl (4-Chloro-7-nitrobenzofurazan, Sigma-Aldrich) as previously described in ref. ²¹. Prior to labeling, free amino acids and other low-molecular-weight compounds were removed using a PD SpinTrap G-25 column (Cytiva).

Proteasome activity was determined in leaves of six-week-old soil-grown plants with the proteasome substrate I (Z-Leu-Leu-Leu-AMC, Sigma-Aldrich) as previously described in ref. ²².

For the darkness treatment and survival rate analysis, plants were germinated on soil and grown under short-day conditions for two weeks. After that, every fourth plant was picked to a pot filled with soil. After one week of recovery growth, plants were moved to a chamber set to night conditions (same conditions as in the short-day chamber, but no light) for 8–14 days. Then the plants were moved back to the short-day conditions for 7–14 days for recovery. The survival rate was calculated by counting the plants with clearly visible intact green meristems and new leaves.

For nitrogen and sulfur starvation assays, Arabidopsis seeds were surface sterilized and seeded on solid ½ Hoagland medium plates (full, -N and -S). Plates were then kept at 4 °C in dark for two days before being moved to LD conditions. After two weeks of growth, the fresh weight of every tenth seedling was weighed. Four replicates were calculated for each phenotype or condition.

The GFP-AtATG8a cleavage assay was performed as described previously⁶⁵. Briefly, five-day-old seedlings grown on ½ MS medium plates containing 1% sucrose under long day conditions were exposed to continuous light for 1 day before being subjected to liquid ½ MS-C medium and kept in dark for 0, 4, 8, and 12 h. Then, the seedlings were harvested to isolate total proteins using NuPAGE buffer for immunoblotting analysis with anti-GFP antibody (Santa-Cruz Biotechnology, sc-9996 HRP).

Protein translation rates in leaves of six-week-old soil-grown plants were assessed by monitoring the incorporation of ³⁵S radiolabeled cysteine and methionine using the EasyTag™ EXPRESS³⁵S Protein Labeling Mix (7 mCi, PerkinElmer, Waltham). In vivo protein labeling and quantification of incorporated radioactivity were performed as previously described in ref. ²¹.

Labeling and detection of newly synthesized proteins via L-azidohomoalanine incorporation was performed as previously described²¹. In brief, leaf discs from six-week-old soil-grown wild type and natb mutant were incubated for 3 h in ½ Hoagland medium supplemented with 50 µM azidohomoalanine. The incorporated azidohomoalanine was conjugated to Biotin Azide (Invitrogen) using the Click-iT™ Protein Reaction Buffer Kit (Invitrogen), and labeled proteins were detected with Neutravidin-HRP (1:100,000 dilution, Invitrogen, 31001). DMSO-treated wild-type leaf discs served as negative controls to distinguish endogenous biotinylated proteins.

Enrichment of all type- and K48-polyubiquitinated proteins was performed using UbiQapture-Q matrix kit (Enzo life Sciences) as previously described in ref. ²¹. Proteasome and deubiquitination activity were inhibited in the leaves of wild type and natb plants by floating leaf discs on ½ Hoagland medium supplemented with 50 μM of MG132 (Santa-Cruz Biotechnology) and 20 μM deubiquitinase inhibitor PR-619 (Sigma Aldrich) for 6 h. To quantify the K63-polyubiquitinated proteins, 1 μM ConA (Santa-Cruz Biotechnology) instead of 50 μM MG132 was applied to the ½ Hoagland medium to inhibit autophagy pathway.

Ubiquitome were analyzed as previously described in ref. ²¹. LFQ values were calculated by MaxQuant and used for quantitative data analysis. Heatmap was plotted by https://www.bioinformatics.com.cn↗ (last accessed on 1st April 2025), an online platform for data analysis and visualization.

Proteins were extracted from dried leaves of 6-week-old soil grown plants using lysis buffer containing 75 mM HEPES pH 7.5, 150 mM KCl, 1.5 mM EGTA pH 8.0, 1.5 mM MgCl₂, 10% (v/v) glycerol, 0.075% NP-40 supplemented with 1 mM DTT, PhosSTOP phosphatase inhibitors, and Protease Inhibitor Mix. Proteins were precipitated in acetone and resolved in 8 M urea following a standard in-solution digest protocol as above. Peptides were separated on an EASY-nLC 1200 HPLC system (Thermo Fisher Scientific, Waltham, MA, USA) using a 110 min gradient from 5–60% acetonitrile with 0.1% formic acid and then directly sprayed in an Orbitrap HF-X Mass Spectrometer (ThermoFisher Scientific, Waltham, MA, USA). The mass spectrometer was operated in a data-independent acquisition mode. The MS1 scans were aquired in a mass range of m/z 350–1000, with an Orbitrap resolution of 120,000 and a normalised AGC target of 3e6. MS2 scans were performed at a resolution of 15,000 and a normalised AGC target of 1e6. 100 windows with an isolation width of 5.6 Th (data points per peak = 6) were aquired per cycle. MS raw files were searched against the UniProt FASTA protein database of Arabidopsis thaliana (release April 2023). Proteins quantified in at least 3 out of 4 replicates in either WT or naa20-cr1 mutant were filtered out, log₂-transformed and imputated. Proteins with a fold change (FC) ≥ 1.5 and a Student’s t-test p-value < 0.05 were considered as significantly different. Heatmap was plotted by https://www.bioinformatics.com.cn↗ (last accessed on 1st April 2025), an online platform for data analysis and visualization.

The N-acetylome was analyzed using the SILProNAQ method on proteins extracted from leaves of six-week-old soil-grown plants⁶⁶. In brief, tissues were frozen, ground, and extracted in buffer (50 mM HEPES pH 7.2, 1.5 mM MgCl₂, 1 mM EGTA, 10% glycerol, 1% Triton X-100, 2 mM PMSF, 150 mM NaCl, and protease inhibitors cocktail). Free N-termini were labeled with d₃-acetyl groups, proteins digested with trypsin, and peptides fractionated by SCX chromatography. Enriched N-terminal peptides were analyzed by LC-MS/MS using an LTQ-Orbitrap Velos, and data were processed with Mascot Distiller and the EnCOUNTer tool⁶⁷.

Seeds of GFP-ATG8a in WT, naa20-cr1, atg5-1, naa20-cr1 atg5-1 backgrounds were surface sterilized and seeded on ½ MS medium plates (with 1% sucrose). Plates were then kept at 4 °C in dark for two days before being moved to LD conditions for one week. Subsequently, plates were transferred to continuous light for one day and then immersed in liquid ½ MS medium (without 1% sucrose) with 1 μM ConA for 6 h in the dark. Seedlings treated with 1 μM DMSO under light were used as control. Roots cells in the elongation zone were imaged using the Leica Stellaris 8, HC PL APO CS2, 20x water-immersed objective and software LAS X 4.5.0.25531 (Leica, Wetzlar). The microscope was set to Pinhole 1 AU, Zoom 1.5, as well as detection with bidirectional scan with phase x30, speed 200 Hz and line average 4, line accumulation 1. The laser intensity was between 5 and 15%. Excitation at 489 nm, GFP fluorescence was recorder with HyD S1 detector at EM = 498–540 nm. Autophagic puncta labeled by GFP-AtATG8a in confocal images were identified using the ImageJ. In brief, images were background-subtracted with 25 pixels of rolling ball radius. Each image was subsequently thresholded using the MaxEntropy method and was converted to an 8-bit grayscale image. Threshold values were adjusted according to the puncta signals in original confocal images. The number of puncta in thresholded images was counted by the Analyze Particles function in ImageJ. For all puncta quantification, puncta with the size between 0.10 and 5.00 μm² were counted.

The root cells of stable transgenic NBR1-mCherry-GFP/WT and NBR1-mCherry-GFP/naa20-cr1 seedlings from ½ MS plates were used to detect the subcellular localization of NBR1 using the Leica Stellaris 8, HC PL APO CS2, 40x water-immersed objective and software LAS X 4.5.0.25531 (Leica, Wetzlar).

Data are presented as mean ± SEM. The number of replicates and experimental details are provided in the respective figure legends. Statistical analyses were performed using GraphPad Prism 9. For comparisons between two groups, an unpaired two-tailed Student’s t-test was applied. Statistically significant differences a indicated with by asterisks (*p < 0.05, **p < 0.01). For analyses involving more than two groups, one-way ANOVA followed by a Tukey´s test was used for comparisons.

Sequence data from this article can be found in The Arabidopsis Information Resource (www.arabidopsis.org↗) under the following accession numbers: AT1G03150 (NAA20), AT5G58450 (NAA25), AT4G24690 (NBR1), AT5G61500 (ATG3), AT5G17290 (ATG5), AT5G45900 (ATG7), AT4G21980 (ATG8a), AT2G45170 (ATG8e), AT2G38130 (NAA30), AT5G11340 (NAA50), AT5G16800 (NAA60), AT1G64520 (RPN12a), AT4G11150 (V-ATPase), AT2G36880 (MAT3), AT1G74310 (HSP101), AT4G31300 (PBA1), AT5G09810 (ACT7), AT1G69960 (PP2A), AT3G01090 (KIN10) and AT3G29160 (KIN11), AT4G35770 (DIN1), AT3G47340 (DIN6), AT5G20250 (DIN10), AT5G28770 (bZIP63).

Further information on research design is available in thelinked to this article. Nature Portfolio Reporting Summary

Nature Communications thanks Daniel Gibbs and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

The mass spectrometry proteomics data associated to N-terminomics of naa20 mutant lines have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD066411↗. The mass spectrometry proteomics data associated to proteomic changes of naa20-cr1 mutant have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD066676↗. The mass spectrometry proteomics data associated to determination of polyubiquitinated proteins enriched with UbiQaputre-Q matrix in the wild-type and naa20-cr1 mutant have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD054814↗. Source data are provided with this paper.