The KIN10 overexpression lines (OE-1 and OE-2) and KIN10 RNAi lines (RNAi-1 and RNAi-7) used in this study were in the Arabidopsis Landsberg erecta (Ler) background (Baena-González et al., 2007). The autophagy-related mutants atg5-1 and atg7-3 (Thompson et al., 2005; Lai et al., 2011; Chen et al., 2015; Col-0 ecotype) were backcrossed twice to the Ler wild-type plants to obtain atg5-L and atg7-L plants. The atg5-L and atg7-L mutants were further crossed to the OE-1 line to generate OE-1 atg5-L and OE-1 atg7-L lines. Transgenic lines expressing GFP-ATG8e and YFP-ATG1a, driven by the CaMV 35S promoter, have been previously described (Xiao et al., 2010; Suttangkakul et al., 2011). All Arabidopsis seeds were surface-sterilized with 20% bleach containing 0.1% Tween-20 for 20 min, and then washed 5 times with sterile water. Seeds were sown on Murashige and Skoog (MS) medium, followed by cold treatment in the dark for 3 days. Seven days after germination, seedlings were transplanted into soil and grown in a plant growth room with a16-h-light/8-h-dark cycle at 22°C.

For the carbon starvation treatment, 7-day-old MS-grown seedlings or 4-week-old soil-grown plants were transferred to continuous darkness for the indicated duration followed by recovery under normal growth conditions for 7 days. Samples were collected or photographed at the indicated time points. To calculate the survival rate after darkness, at least 18 plants per genotype were dark-treated followed by a 7-day recovery. The number of surviving plants, where survival if defined as the capability to produce new leaves, was recorded. For the nitrogen starvation treatment, 7-day-old seedlings grown on MS medium were transferred to solid MS or nitrogen-deficient solid MS medium and grown for 5 days. Chlorophyll contents were measured and calculated after the recovery.

Total RNA extraction and quantitative reverse-transcription PCR (qRT-PCR) analysis were performed as previously described (Chen et al., 2015). Briefly, the isolated RNA was reverse transcribed using the PrimeScript RT Reagent Kit with gDNA Eraser (Takara, RR047A) following the manufacturer’s instructions. The qPCR was carried out using SYBR Green master mix (Takara, RR420A) on a StepOne Plus real-time PCR system (Applied Biosystems). The conditions for the qPCR were: initial denaturation at 95°C for 5 min followed by 40 cycles of PCR (denaturing at 95°C for 10 s, annealing at 55°C for 15 s, and extension at 72°C for 30 s). Three experimental replicates were used for each reaction. ACTIN2 was used as the reference gene. Gene-specific primers used for the qPCR analysis are listed in Supplementary Table S1.

The stable transgenic lines expressing a GFP-ATG8e fusion protein were used to monitor autophagosome formation (Xiao et al., 2010). Seven-day-old GFP-ATG8e seedlings grown in MS solid medium were transferred to MS medium or MS medium lacking sugars (MS-C) under darkness for the indicated times. After treatment, the primary root cells were observed using an LSM 780 NLO laser scanning confocal microscope (Carl Zeiss).

Total protein extraction was performed as previously described (Chen et al., 2015). Briefly, 4-week-old plant leaves or 7-day-old whole seedling were ground in liquid nitrogen and homogenized in ice-cold extraction buffer (50 mM sodium phosphate, pH 7.0, 200 mM NaCl, 10 mM MgCl₂, 0.2% β-mercaptoethanol and 10% glycerol) supplemented with protease inhibitor cocktail (Roche, 04693132001). Total homogenates were placed on ice for 30 min, and then centrifuged for 30 min at 11,000 g. The supernatant was transferred to a new microfuge tube for SDS-PAGE electrophoresis.

For immunoblot analysis, total proteins were subjected to SDS-PAGE and electrophoretically transferred to a Hybond-C membrane (Amersham, 10600016). Anti-GFP antibodies were used to detect GFP as previously described (Chen et al., 2015). YFP was detected with rabbit anti-GFP polyclonal antibodies (Abcam, ab290).

Phosphatase treatment was performed according to Suttangkakul et al. (2011) with minor modification. Seven-day-old YFP-ATG1a and YFP-ATG1a/KIN10-OE seedling were homogenized in ice-cold protein extraction buffer supplemented with 1 mM phenylmethylsulfonyl fluoride and protease inhibitor cocktail (Roche). Samples were placed on ice for 30 min, and then centrifuged for 30 min at 11,000 g. The supernatant was incubated with λ protein phosphatase (New England Biolabs) with or without addition of phosphatase inhibitor PhosSTOP (Roche) for 30 min at 30°C. 2 × SDS-PAGE sample buffer was added to the total sample and heated to 95°C for 5 min.

Data are reported as means ± SD of three independent experiments unless otherwise indicated. The significance of the differences between groups was determined by a two-tailed Student’s t-test. P-values < 0.05 or < 0.01 were considered significant.

Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank databases under the following accession numbers: KIN10 (At3g01090), ATG1a (At3g61960), ATG1b (At3g53930), ATG1c (At2g37840), ATG2 (At3g19190), ATG5 (At5g17290), ATG6 (At3g61710), ATG7 (At5g45900), ATG8a (At4g21980), ATG8e (At2g45170), ATG9 (At2g31260), ATG10 (At3g07525), ATG13a (At3g49590), ATG13b (At3g18770), ATG18a (At3g62770), and PI3K (At1g60490).

The KIN10 overexpression lines (OE-1 and OE-2) showed delayed leaf senescence (Baena-González et al., 2007). To examine the potential role of Arabidopsis KIN10 in autophagy, we further examined the response of the KIN10-OE lines and KIN10 RNA interference lines (RNAi-1 and RNAi-7) to naturally induced senescence and nutrient deficiency. The RNA and protein level of KIN10 in the KIN10-OE and KIN10-RNAi lines were first confirmed by qRT-PCR and western blot analyses (Supplementary Figure S1). Under normal growth conditions, both KIN10-OE lines displayed slower growth and delayed natural senescence compared to the wild type, while the KIN10-RNAi lines showed similar phenotypes to the wild-type plants (Figure 1A). The level of chlorophyll in the leaves of 6-week-old KIN10-OE lines was much higher than that of the wild type (Figure 1B).

The KIN10-OE lines showed enhanced tolerance to carbon starvation induced by constant darkness for 7 days followed by a 7-day recovery, while the KIN10-RNAi lines appeared similar to the wild-type plants after this treatment (Figures 1C,D). For the nitrogen deficiency treatment, 7-day-old MS-grown seedlings were transferred to MS or MS-N solid medium for 5 days. The cotyledons of the wild-type plants and KIN10-RNAi lines were significantly yellowed as indicated by the reduced chlorophyll contents (Figures 1E,F). In contrast, the KIN10-OE lines were more resistant to nitrogen deficiency and had significantly higher chlorophyll levels compared to that of the wild type (Figures 1E,F). These findings suggest that overexpression of KIN10 can delay natural senescence and improves tolerance to carbon and nitrogen starvation.

Given that the KIN10-overexpression lines showed delayed leaf senescence and enhanced tolerance to nutrient starvation (Figures 1, 2), and that the autophagy-deficient mutants had the opposite phenotype (Baena-González et al., 2007; Li and Vierstra, 2012), we used these plants to further assess the genetic connection between KIN10 and the autophagy pathway. The atg5-L and atg7-L mutants (atg5 and atg7 mutants in the Ler background) were crossed to KIN10-OE-1 (OE-1) to generate OE-1 atg5-L and OE-1 atg7-L lines. We then tested the tolerance of the 4-week-old wild-type, OE-1, OE-1 atg5-L, OE-1 atg7-L, atg5-L, and atg7-L plants to carbon starvation. Compared to the wild-type plants, the OE-1 plants showed enhanced tolerance and the atg5-L and atg7-L plants showed decreased tolerance to carbon starvation. Interestingly, the OE-1 atg5-L and OE-1 atg7-L lines displayed similar sensitivities to carbon starvation to that of the atg5-L and atg7-L mutants (Figure 2A). In addition, the enhanced resistance of the OE-1 line to nitrogen deficiency was attenuated by the loss-of-function of ATG5 and ATG7 (Figure 2B). The enhanced tolerance to starvation in the OE-1 line was further supported by the higher survival rates (Figure 2C) and higher relative chlorophyll contents (Figure 2D) in this line. Together, these results indicate that the enhanced tolerance to nutrient starvation in the KIN10-OE lines is dependent on a functional autophagy pathway. The evidence that the autophagy-associated phenotypes in the KIN10-OEs were primarily recovered by the autophagy deficient mutants, suggesting that autophagy acts downstream of KIN10 to affect plant growth and stress responses. Given that KIN10 is a well-known master regulator in energy signaling in Arabidopsis, we therefore proposed that it governs a cellular switch between plant growth and stress responses by modulating various downstream signaling pathways, including autophagy.

The autophagy-defective mutants are hypersensitive to abiotic stresses such as drought and submergence (Liu et al., 2009; Chen et al., 2015). To further assess the role of KIN10 in autophagy-related stress responses, the wild-type (Ler), the KIN10-OE lines (OE-1 and OE-2), and the KIN10-RNAi lines (RNAi-1 and RNAi-7) were subjected to drought and submergence treatments. As shown in Figure 3, no significant differences were observed between the wild type and the KIN10-OE or KIN10-RNAi lines under normal growth conditions. However, after a 14-day drought treatment, the leaves of the wild type and the KIN10-RNAi lines turned yellow and wilted, while the leaves of the KIN10-OE lines remained green and turgid (Figure 3A). After a 4-day recovery by rehydration, the survival rates of the KIN10-OE lines were significantly higher than those of the wild type and the KIN10-RNAi lines (Figure 3B). In addition, the KIN10-OE lines were much more tolerant than the wild type and the KIN10-RNAi plants to a 6-day submergence in water (under light conditions) followed by a 6-day recovery (Figure 3C), which was supported by the improved survival rates of the KIN10-OE lines compared to the wild-type plants after submergence (Figure 3D).

To examine the potential involvement of KIN10 in regulating autophagosome formation, we first tested the abundance of ATG transcripts (ATG2, ATG5, ATG7, ATG8a, ATG10, and ATG18a) in the wild type, KIN10-OE lines (OE-1 and OE-2), and the KIN10-RNAi lines (RNAi-1 and RNAi-7). qRT-PCR analyses showed no significant changes in the expression levels of ATG7, ATG10, and ATG18a among the wild type, KIN10-OE lines, or the KIN10-RNAi lines, while the expression of ATG2, ATG5, and ATG8a was slightly upregulated in the KIN10-OE lines in comparison to the wild type (Supplementary Figure S2).

To further investigate the role of KIN10 in the induction of autophagy, we examined autophagosome formation in the wild type, KIN10-OE and KIN10-RNAi lines using green fluorescent protein (GFP)-tagged ATG8e (Xiao et al., 2010). Seven-day-old GFP-ATG8e (wild-type background), GFP-ATG8e/KIN10-OE and GFP-ATG8e/KIN10-RNAi seedlings were transferred to solid MS medium (MS) or MS-C under darkness for 6 h, and the GFP fluorescence of root cells was subsequently observed using confocal microscopy. As shown in Figure 4, under both MS and MS-C conditions, the numbers of GFP-ATG8e labeled punctate structures significantly increased in the KIN10-OE lines in comparison to the wild type and the KIN10-RNAi lines (Figures 4A,B). Upon nutrient starvation, the GFP-ATG8e fusion protein is degraded to release a free, relatively stable GFP, and the accumulation of GFP signals reflects the level of induction of autophagy (Li et al., 2014; Klionsky et al., 2016). As shown in Figure 4C, the degradation of the GFP-ATG8e fusion protein in the GFP-ATG8e/KIN10-OE line was faster than in the GFP-ATG8e or GFP-ATG8e/KIN10-RNAi line. Consistent with this, the ratio of free GFP to GFP-ATG8e in the GFP-ATG8e/KIN10-OE line was higher than that of in the GFP-ATG8e or GFP-ATG8e/KIN10-RNAi line (Figure 4D), suggesting that overexpression of KIN10 enhances autophagic activity.

Given that AMPK phosphorylates ULK1 to activate autophagy in mammalian cells (Egan et al., 2011; Kim et al., 2011), we hypothesized that KIN10 may be involved in autophagy by directly or indirectly phosphorylating ATG1. To confirm this, ATG1 phosphorylation was first tested using λ protein phosphatase and phosphatase inhibitor PhosSTOP in a yellow fluorescent protein (YFP)-tagged ATG1a transgenic plant (YFP-ATG1a) (Suttangkakul et al., 2011). As shown in Figure 5A, two species of YFP-ATG1a were detected using anti-GFP antibodies by western blot. Consistent with a previous study (Suttangkakul et al., 2011), the λ phosphatase treatment of total protein extracted from the YFP-ATG1a transgenic plant reduced the levels of the higher molecular weight species, while PhosSTOP blocked this shift (Figure 5A), To demonstrate the role of KIN10 in the regulation of ATG1, we crossed the YFP-ATG1a line to the OE-1 line to generate the YFP-ATG1a/KIN10-OE lines. Immunoblot analysis showed that the level of YFP-ATG1a fusion protein was significantly higher in the YFP-ATG1a KIN10-OE line than in the YFP-ATG1a line (Figure 5B). Upon carbon starvation, the phosphorylation status of ATG1 was enhanced in the YFP-ATG1a KIN10-OE line in comparison to the YFP-ATG1a line (Figures 5B,C).

To determine whether the accumulation of YFP-ATG1a was caused by the higher transcription of YFP-ATG1a in the YFP-ATG1a KIN10-OE line, we analyzed the total transcript level of ATG1a and YFP-ATG1a during carbon starvation by qRT-PCR. As shown in Supplementary Figure S3, The total transcript level of ATG1a was enhanced in the YFP-ATG1a KIN10-OE line but not in the YFP-ATG1a line, while the YFP-ATG1a transcript level did not change much in either line. Interestingly, the total expression of both ATG1a and YFP-ATG1a was slightly higher in the YFP-ATG1a KIN10-OE line than in the YFP-ATG1a line.