SNF1‐related protein kinase 1 represses Arabidopsis growth through post‐translational modification of E2Fa in response to energy stress

Summary Cellular sugar starvation and/or energy deprivation serves as an important signaling cue for the live cells to trigger the necessary stress adaptation response. When exposed to cellular energy stress (ES) conditions, the plants reconfigure metabolic pathways and rebalance energy status while restricting vegetative organ growth. Despite the vital importance of this ES‐induced growth restriction, the regulatory mechanism underlying the response remains largely elusive in plants. Using plant cell‐ and whole plant‐based functional analyses coupled with extended genetic validation, we show that cellular ES‐activated SNF1‐related protein kinase 1 (SnRK1.1) directly interacts with and phosphorylates E2Fa transcription factor, a critical cell cycle regulator. Phosphorylation of E2Fa by SnRK1.1 leads to its proteasome‐mediated protein degradation, resulting in S‐phase repression and organ growth restriction. Our findings show that ES‐dependently activated SnRK1.1 adjusts cell proliferation and vegetative growth for plants to cope with constantly fluctuating environments.


Introduction
Plants have integrated developmental, metabolic, and physiological responses to survive unfavorable growth conditions (e.g.abiotic stress, limited light, or nutrient).Cellular energy stress (ES) conditions serve as a signaling cue to trigger such stress adaptation processes, leading to metabolite redistribution to quiescent meristems and/or storage organs while restraining vegetative organ growth (Rolland et al., 2006;Zhu, 2016;Crepin & Rolland, 2019).For example, most land plants grow slowly and eventually terminate vegetative growth under low oxygen and extended darkness (Cho et al., 2012;Lim et al., 2021).Despite the vital importance of the plant's ability to adapt and sustain under cellular ES conditions, cellular signaling and regulatory mechanisms in the stress adaptation process remain largely elusive.
Cell proliferation is a key determinant of organismal growth, development, and patterning (Komaki & Sugimoto, 2012;Scofield et al., 2014).Regulatory modules for cell cycle-dependent proliferation are widely conserved in most eukaryotes, including plants.Evolutionarily conserved genes encoding for cell cycle regulators are found in a diverse array of plant genomes (Harashima et al., 2013;Liu et al., 2018).The Arabidopsis thaliana genome contains several key transcription factors of cell cycle regulatory modules (e.g.E2Fa, E2Fb, and E2Fc) that have high sequence similarities to those in many other species (Vandepoele et al., 2002;Magyar et al., 2005).E2Fa and E2Fb are transcription activators responsible for S-phase-specific gene expression, contributing to DNA replication, DNA repair, and chromatin maintenance (Ramirez-Parra et al., 2003;Vandepoele et al., 2005;Takahashi et al., 2008;Leviczky et al., 2019).In addition, certain plant hormones have been known for their function in the G1-to-S transition during plant cell division.For example, the plant stress hormone abscisic acid (ABA) reduces CDKA (cyclin-dependent kinase A) level (Garza-Aguilar et al., 2017), while another plant stress and defense hormone, ethylene, suppresses CDKA activity, resulting in plant growth suppression (Skirycz et al., 2011).
Under cellular energy starvation conditions, phosphorylation of the C-group bZIP transcription factor bZIP63 by SnRK1.1 plays a role in the transcriptional reconfiguration of metabolic pathway and reduces dark-induced senescence (Mair et al., 2015).Nuclear localization of SnRK1.1 is rhythmic and plays a role in not only metabolic stress responses but also organ growth and development (Yuan et al., 2016;Ramon et al., 2019).Thus, a better understanding of the nuclear activity of SnRK1.1 promises to yield deeper insights into how the ES-inducible protein kinase activity would contribute to life-sustaining processes in plant cells under unfavorable environmental conditions.
Here, we report that SnRK1.1 directly interacts with and phosphorylates the marked box (MB) domain (Mariconti et al., 2002) of the transcription factor E2Fa.The SnRK1.1-mediated phosphorylation of E2Fa leads to proteasome-dependent protein degradation of E2Fa, resulting in downregulation of S-phasespecific gene expression and cell cycle arrest under ES conditions.Our findings reveal that a cellular energy sensor SnRK1.1 serves as a key regulatory element in the control of cell proliferation for plant survival and adaptation to unfavorable growth conditions.
For e2fa complementary line generation, the promoter of E2Fa (pE2Fa) was PCR-amplified and substituted with the 35S promoter in the pBI121 binary vector (pE2Fa_pBI121).A full-length E2Fa cDNA was PCR-amplified from Col-0 cDNA and cloned into the C-terminal 2xMYC-tagged HBT vector.The E2Fa-2xMYC2-NOS terminator cassette from the HBT vector was fused to the pE2Fa_pBI121 vector by restriction enzyme digestion and ligation (pE2Fa::E2Fa-2xMYC2).The E2Fa T314AT315A was generated by site-direct mutagenesis and cloned into the 2xHA-tagged HBT vector.The E2Fa T314AT315A -2xHA-NOS terminator cassette from the HBT vector was fused to the pE2Fa_pBI121 vector by restriction enzyme digestion and ligation.These constructs were transformed into the e2fa mutant line.The transgenic plants were selected as single-insertion homozygous lines.
Arabidopsis seeds were grown either in six-well plates (six seeds per well) containing 1 ml of ½ Murashige & Skoog (½MS) liquid medium (0.5% sucrose, pH 5.7 adjusted with KOH) or in soil for indicated days in a growth room at 23°C, 50% humidity, and 75 lmol m À2 s À1 light intensity under 16 h : 8 h, light : dark photoperiod.

Energy stress and phytohormone treatments
Four-day-old Arabidopsis seedlings were transferred to a hypoxic chamber with 1% O 2 , < 0.1% CO 2 , and 98.9% N 2 (Ruskinns Invivo 300 ; Ruskinn Technology Ltd, Bridgend, UK) in submerged and dark conditions with and without carbon supplementation.For phytohormone treatment, 4-d-old seedlings were transferred to ½MS liquid medium (0.5% sucrose, pH 5.7 adjusted with KOH) containing 100 lM ABA, ACC, or MeJA, and then incubated for 6 h.

Transient promoter assay in mesophyll protoplasts
Arabidopsis mesophyll protoplast transient promoter assays were performed as described previously (Yoo et al., 2007).To generate effector constructs, full-length cDNA of SnRK1.1 WT , SnRK1.1 IN , E2Fa, E2Fb, and E2Fc was PCR-amplified and cloned into the 35SC4PPDK promoter and the NOS terminator in plant cell expression vector, pHBT.E2Fa S93A , E2Fa T99A , E2Fa S170A , E2Fa S278A , E2Fa T314AT315A , E2Fa S427A , and E2Fa S454A were generated with site-direct mutagenesis.SPYNE (the N-terminal end of split YFP) was used as effector control.To generate promoter reporter, the 35SC4PPDK promoter of pHBT-fLUC (firefly luciferase) was replaced by target promoter.For protein translation reporter, full-length cDNA of target protein cloned into the 35SC4PPDK promoter and fLUC in pHBT-fLUC.The 35S-driven rLUC (renilla luciferase) in pHBT was used as an expression control.A total of 40 lg of the desired DNA constructs were transfected to protoplasts (6 9 10 4 cells per 200 ll) by the PEG method and incubated in WI solution for 6 h at 25°C.For chemical treatment, DMSO or 5 lM MG132 was added to the solution after 1 h of incubation.The experiments were independently carried out at least three times, and the data were analyzed by t-test.

Co-immunoprecipitation
Co-immunoprecipitation (Co-IP) was performed as described previously (Xiong et al., 2013) with slight modifications.Briefly, 300 lg of each E2Fa-GFP was transfected with or without 100 lg of SnRK1.1-HA to 2 ml of protoplasts (4.5 9 10 5 cells) and incubated 11 h in W5 solution containing 5 lM MG132.After washing with WI solution, transfected protoplasts were lysed in 200 ll of lysis buffer.The samples were then spun down, and the supernatant was transferred into 2-ml tubes.After the transfer of 20 ll of the sample as an input fraction, the volume of the lysate-supernatants was adjusted to 500 ll with dilution buffer.Twenty-five microliters of bead slurry (ChromoTek, Planegg, Germany) was added and incubated for 6 h on shaker at a 4°C.After the beads were washed three times with 500 ll ice-cold wash buffer, the samples were resuspended with 50 ll of 29 sample buffer.The presence of co-immunoprecipitated proteins was detected by immunoblotting.

Gene expression analysis
To analyze gene expression, total RNA was isolated from plants using the TRIzol reagent (Invitrogen), and 1 lg of total RNA was used for cDNA synthesis using M-MLV reverse transcriptase (Promega).Reverse transcription polymerase chain reaction was performed using CFX96 Real-Time System (Bio-Rad) with gene-specific primers listed in Table S1.The UBQ10 and EIF4a genes were used as expression controls.
New Phytologist (2023) 237: 823-839 www.newphytologist.com 4 mM PMSF) and centrifuged twice at 16 000 g for 15 min at 4°C, and the supernatants were collected into new tubes.After protein concentrations were measured by the Bio-Rad protein assay, total protein extracts from Col-0 and SnRK1.1 WT transgenic plants were adjusted to equal concentration with protein stability assay buffer.Purified, recombinant GST-E2Fa and GST-E2Fa T314AT315A proteins generated in E. coli were incubated for indicated time in equal quantities of plant extracts with DMSO or 40 lM MG132 at 22°C, and the reaction was stopped by 59 SDS-PAGE sample buffer.An equal volume of each sample was separated in 8% SDS-PAGE gel, and immunoblots were performed.

Protein mobility shift assay
Protein mobility shift assay was performed as described previously with slight modification (Im et al., 2014).For the protein mobility shift assay, E2Fa-MYC and E2Fa T314AT315A -MYC were expressed in Arabidopsis protoplasts with 10 lM MG132 with or without SnRK1.1 WT , then separated on a 10% (w/v) SDS-PAGE for 2 h, and identified by protein blot analysis using anti-MYC antibody.

In-gel digestion
In-gel digestion was performed using a modified protocol of a previously reported method (Shevchenko et al., 2007).In vitro kinase assay was carried out with SnRK1.1 and E2Fa and separated the proteins with PAGE.A band of E2Fa was sliced into small pieces (c. 1 9 1 mm 2 ) with a clean blade, and the gel pieces were transferred into a microcentrifuge tube.The samples were used for the analysis of the phosphorylation site of E2Fa by LC-MS/MS.

Statistical analysis
All experiments were independently conducted at least three times, and the data were analyzed by t-test or ANOVA.Asterisks denote significant differences (**, P < 0.01; *, P < 0.05), and different letters indicate statistical differences (P < 0.05).

Cellular energy deprivation induces organ growth restriction
With the cellular ES treatment of low oxygen (< 1%) and extended darkness that interfere normal photosynthesis (Fig. S1a), the growth of rosettes, cotyledons, and primary roots was reduced (Fig. S1b,c).The ES-induced growth reduction was reversed when sucrose was added to the medium (Fig. S1c).
To examine whether the cellular ES-induced growth restriction was due to a lack of cell division, S-phase cell division activity in the primary root meristems was monitored by EdU staining, which detects DNA synthesis in proliferating cells (Xiong et al., 2013).EdUfluorescence was detected in the cell division zone of the primary roots but was abolished following ES treatment (Fig. S1d), indicating that cell division was stopped with the ES treatment.
We then examined the expression of several marker genes related to cellular ES and S-phase activity in Arabidopsis seedlings.The expression of two SnRK1.1-activatedgenes DARK INDUCIBLE 1 (DIN1) and DIN6 (Baena- Gonzalez et al., 2007) was significantly increased by the ES treatment (Fig. S1e), while the expression of two key S-phase-specific transcription activators E2Fa and E2Fb was slightly reduced to about 70-80% by the ES treatment, compared with the control plants (Fig. S1f).However, the expression of the direct target genes of E2Fs was decreased markedly (Fig. S1f), suggesting that ES restricts plant growth through negative regulation of S-phase activity.

SnRK1.1 is activated by energy stress and restricts organ growth
Given that SnRK1 plays a fundamental role in plant response to cellular ES (Hulsmans et al., 2016) and the expression of SnRK1.1-inducibleDIN1 and DIN6 was significantly increased by ES treatment (Fig. S1e), we reasoned that ES may activate SnRK1.1 kinase activity.To test this hypothesis, we first analyzed SnRK1.1 protein accumulation using anti-SnRK1.1 antibodies and then measured T-loop-phosphorylated SnRK1.1 (P-SnRK1.1;i.e. its activation) using anti-pT172-AMPKa antibodies (Crozet et al., 2014).Both SnRK1.1 and P-SnRK1.1 protein levels were increased compared with those of Actin11 by the cellular ES treatment (Fig. 1a,b).
We further hypothesized that increased activity of SnRK1.1 may lead to growth restriction through the inhibition of S-phase.To test this hypothesis, we used transgenic Arabidopsis plants overexpressing either hemagglutinin (HA) epitope-tagged normal SnRK1.1 (SnRK1.1 WT ) or inactive form of SnRK1.1 (SnRK1.1 IN ).Since single-knockout mutants of SnRK1s such as snrk1.1 and snrk1.2did not produce any detectible phenotype, we used SnRK1.1 IN as a loss-of-function variant in this study (Fig. S2).The SnRK1.1 IN was created by substituting Lys48 with Met residue to block the ATP-binding site of the protein kinase (Baena- Gonzalez et al., 2007;Im et al., 2014).While both SnRK1.1 WT -HA and SnRK1.1 IN -HA proteins were clearly detected using anti-HA antibody in the respective transgenic plants, the level of T-loop phosphorylated SnRK1.1 was substantially reduced in the SnRK1.1 IN plants (Fig. 1c), suggesting that SnRK1.1 IN may function as dominant-negative regulator in the SnRK1 signaling as reported previously (Baena- Gonzalez et al., 2007).
As expected, SnRK1.1 WT transgenic plants had smaller rosettes than did WT Col-0 and SnRK1.1 IN transgenic plants (Fig. 1d).Also, the SnRK1.1 WT transgenic seedlings had shorter primary roots than the Col-0 plants (Fig. 1e), suggesting that SnRK1.1 may negatively regulate plant growth.To corroborate this observation, we analyzed DNA synthesis (i.e.cell division) activity in the primary roots using Edu assay.Edu staining was much lower in the SnRK1.1 WT plants than in Col-0 and SnRK1.1 IN transgenic seedlings (Fig. 1f).
To further confirm the SnRK1.1-mediatedcell division repression, the expression of S-phase-specific genes E2F TARGET

SnRK1.1 represses G1-S transition
Given that SnRK1.1 negatively regulates S-phase activity and is known to regulate transcription factors, we asked whether a key cell cycle regulator E2Fa is involved in the SnRK1.1-mediatedSphase suppression.To genetically test this, we investigated the growth of cotyledons and primary roots, which are under the control of cell division, in various SnRK1.1 and E2Fa lines (Cho et al., 2012;Magyar et al., 2012;Xiong et al., 2013).The SnRK1.1 WT , e2fa, and E2Fa RNAi transgenic plants had shorter primary root length and smaller cotyledon than the WT Col-0 plants (Fig. S3a).However, the cotyledons and primary roots of  Gonzalez et al., 2007;Cho et al., 2012), we asked whether E2Fa is also involved in the SnRK1.1-mediatedcontrol of inflorescence development and senescence.While E2Fa positively regulated both cotyledon size and root length, it did not have an impact on the development of inflorescence and senescence (Fig. S3b), suggesting E2Fa is involved in SnRK1.1 signaling related to plant growth but not plant development.

E2Fa protein is negatively regulated under energy stress condition
Based on the observation that the ES treatment eliminated the expression of E2Fa target genes, albeit the expression of E2Fa gene itself was very slightly reduced by the treatment (Fig. S1f), we hypothesized that E2Fa activity might be post-translationally regulated under ES conditions.To test this hypothesis, we investigated E2Fa protein stability using transgenic plants expressing green fluorescence protein (GFP) fused to the C-terminus of E2Fa gene under the control of own promoter (pE2Fa::gE2Fa-GFP).
With ES treatment, E2Fa-GFP protein was rapidly decreased while P-SnRK1.1 and SnRK1.1 protein levels were increased (Fig. 2a).Consistent with the observed instability of E2Fa protein, the expression of its direct target genes (ETG1, MCM5, ORC2, and PCNA1) started to decrease after 6 h of the ES treatment (Fig. 2b), suggesting that ES may negatively regulate E2Fa transcriptional activity.

New Phytologist
To investigate whether stress-related hormone signaling is involved in the ES-mediated degradation of E2Fa protein, we analyzed E2Fa protein stability in response to ES and plant stress hormone treatments (ABA, ACC, and MeJA).The hormone treatments were applied for 6 h, at which time point most of E2Fa protein is degraded and its target gene expression diminishes by ES treatment (Fig. 2a,b).The fluorescence signal of the E2Fa-GFP was abolished by ES, but not by the stress hormones (Fig. 2c).Likewise, E2Fa protein level was reduced significantly by ES, while no change was observed by ABA, ACC, or MeJA treatment (Fig. 2d).These results suggest that, unlike ES, the stress-related hormones do not affect E2Fa protein level.

SnRK1.1 suppresses E2Fa-and E2Fb-inducible S-phasespecific gene expression
As a part of the functional analysis of SnRK1.1 kinase in cell cycle-related gene regulation, we used fLUC reporter gene under the control of either S-phase-specific ETG1 or MCM5 promoter (pETG1-fLUC or pMCM5-fLUC) to analyze E2Fa-, E2Fb-, and E2Fc-dependent reporter fLUC activities in Arabidopsis protoplasts.These reporter constructs were transfected to protoplasts with E2Fs, SnRK1.1 WT , or SnRK1.1 IN as effector.The promoter activity was increased with E2Fa or E2Fb expression but repressed significantly by SnRK1.1 WT co-expression but not by SnRK1.1 IN (Fig. 3).However, E2Fc, which is known as transcription repressor, did not affect the activity of the S-phase reporter (Fig. 3).Taken together, SnRK1.1 kinase activity effectively suppresses E2Fa-and E2Fb-mediated gene expression.

SnRK1.1 directly interacts with and phosphorylates E2Fa
To evaluate the nature of SnRK1.1 and E2Fa/b interaction, we carried out yeast two-hybrid (Y2H) analyses that showed an interaction between SnRK1.1 and E2Fa or E2Fb (Fig. S4a).Moreover, the C-terminal regulatory domain of SnRK1.1 interacted with E2Fa or E2Fb (Fig. S4b), indicating that SnRK1.1 may modulate E2Fa and E2Fb functions through direct proteinprotein interaction.We then examined whether SnRK1.1 phosphorylates E2Fa and E2Fb by carrying out in vitro SnRK1.1 kinase assay using glutathione S-transferase (GST)-tagged E2Fa/ b as protein substrates.E2Fc was used as a negative control substrate.GST-SnRK1.1 and its upstream activator GST-GRIK1were used in the in vitro kinase assay.The assay showed that GRIK1-activated SnRK1.1 could phosphorylate E2Fa and E2Fb but not E2Fc (Fig. S4c), confirming that SnRK1.1 phosphorylates the key cell cycle transcription activator E2Fa and E2Fb.Although E2Fa and E2Fb appear to be redundantly required for cell cycle progression, E2Fa is expressed predominantly in primary roots, which largely overlaps with SnRK1.1 expression (Fig. S4d).For this reason, we focused on SnRK1.1-E2Faregulon.
To further ascertain the protein-protein interaction between SnRK1.1 and E2Fa in vivo, we carried out Co-IP analyses in Arabidopsis protoplasts.HA-tagged SnRK1.1 was detected in the immunoprecipitated complex with E2Fa-GFP, but not in the control (Figs 4a, S5), confirming that SnRK1.1 interacts with E2Fa in vivo.To gain additional insights into the regulatory mechanism involved in the SnRK1.1-dependentE2Fa repression, we investigated SnRK1.1-bindingregion(s) in E2Fa.The Y2H assay showed that the E2Fa 282-352 fragment including MB domain, a potential heterodimerization and DNA bending motif (Mariconti et al., 2002), is responsible for the SnRK1.1 interaction (Fig. 4b).
Since SnRK1.1 phosphorylated E2FaMB T314A relatively weakly (Fig. 4e) and MS analysis identified only T315 as a phosphorylation site (Fig. 4d), we hypothesized that T314 and T315 might be phosphorylated in a sequential manner.Furthermore, since our bioinformatics analysis identified not only T314 and T315 but also T339 as the potential SnRK1.1-dependentphosphorylation site (Fig. S6a), their phosphorylation could perhaps lead to subsequent phosphorylation of S338 or T339.To test this hypothesis, we examined whether E2FaMB T315D , which mimics phosphorylation at T315, could promote subsequent phosphorylation at other sites.SnRK1.1 phosphorylated E2FaMB T315D to a certain level, but not E2FaMB T314AT315D (Fig. 4f).By contrast, SnRK1.1 did not phosphorylate E2FaMB T314DT315D , E2FaMB T314DT315DS338A , or E2FaMB T314DT315DT339A .These results indicate that SnRK1.1 phosphorylates T315 and T314 in the E2Fa MB domain, but their phosphorylation does not result in subsequent phosphorylation of S338 and T339.

SnRK1.1 negatively modulates E2Fa and E2Fb protein stability through 26S proteasome-mediated protein degradation
In light of the findings that E2Fa activity is post-translationally regulated (Figs 2, 3) and SnRK1.1 phosphorylates E2Fa (Fig. 4), we analyzed E2Fa and E2Fb protein levels with co-expression of either SnRK1.1 WT or SnRK1.1 IN .Both E2Fa and E2Fb protein levels were reduced in the Arabidopsis protoplasts co-expressing SnRK1.1 WT but not by SnRK1.1 IN co-expression (Fig. 5a).Consistent with the protein blot analysis results, the fluorescence signal of GFP-tagged E2Fa and E2Fb in the nucleus was also compromised by SnRK1.1 WT co-expression (Fig. 5b).To further confirm these findings, we used protein translation reporters constructed by translational fusion of either E2Fa or E2Fb cDNA to fLUC reporter gene under the control of 35SC4PPDK promoter.The fLUC activity of the protein translation reporters reflects the protein accumulation levels of the fusion proteins.First, E2Fa-fLUC and E2Fb-fLUC activities were significantly reduced by SnRK1.1 co-expression, while E2Fa-fLUC and E2Fb-fLUC activities were maintained in the presence of the proteasome inhibitor MG132 (Fig. 5c), suggesting that SnRK1.1 negatively modulates E2Fa and E2Fb protein stability through 26S proteasome-mediated protein degradation.
In cell cycle regulation, RETINOBLASTOMA-RELATED-PROTEIN (RBR) negatively regulates E2F-dependent activation of S-phase genes (Harashima & Sugimoto, 2016).We asked whether SnRK1.1 activity affects RBR-dependent regulation of E2Fa/b function.Fig. 5d shows that SnRK1.1-mediatedsuppression of E2Fa/b activity was not altered by RBR co-expression, indicating that SnRK1.1 and RBR may repress E2Fa/b in different manners.Moreover, RBR did not affect E2Fa/b protein accumulations, while SnRK1.1 reduced E2Fa/b protein levels regardless of RBR co-expression (Fig. 5d).These results suggest that SnRK1.1 suppresses E2Fa/b functions through their protein degradation, which is distinct from the RBR-dependent negative transcriptional regulation of E2F during cell cycle progression.
We then hypothesized that phosphorylation of T314 and T315 of E2Fa by SnRK1.1 leads to phosphorylation-dependent degradation of the protein.To test this hypothesis, we carried out cell-free protein turnover assay for GST-E2Fa and GST-E2Fa T314AT315A .Protein abundance was measured for GST-E2Fa or GST-E2Fa T314AT315A incubated in the protein extracts isolated from WT Col-0 or SnRK1.1 WT transgenic plants.GST-E2Fa decreased faster in the protein extracts of SnRK1.1 WT transgenic plants than in the extracts of Col-0 (Fig. 7c).On the contrary, E2Fa protein level maintained in the presence of protease inhibitor MG132, verifying 26S proteasome-mediated degradation of E2Fa.As predicted, degradation of GST-E2Fa T314AT315A was slower than that of GST-E2Fa in the protein extracts isolated from SnRK1.1 WT transgenic plants (Fig. 7d).
To further confirm that E2Fa phosphorylation regulates the protein stability of E2Fa in planta during ES, we generated transgenic plants expressing pE2Fa::E2Fa-MYC or pE2Fa:: E2Fa T314AT315A -HA in e2fa mutant background.Multiple homozygous single transgene insertion lines with high level of transgene expression were selected and used in the analysis (Fig. S9).Consistent with the results from in vitro assay of protein turnover, E2Fa T314AT315A -MYC protein accumulation was higher than that of E2Fa-MYC in e2fa complemented transgenic plants under ES conditions (Fig. 7e).The expression of E2Fa target genes also was significantly higher in pE2Fa::E2Fa T314AT315A -HA/e2fa than in pE2Fa::E2Fa-MYC/e2fa under ES conditions (Fig. 7f).These data indicate that T314/T315 residues of E2Fa contribute to the protein degradation of E2Fa in plant under ES conditions.
The two phosphorylation sites of E2Fa are critical for the ES-induced root growth restriction With ES treatment, primary roots of pE2Fa::E2Fa T314AT315A -HA/e2fa plants grew longer than those of pE2Fa::E2Fa-MYC/e2fa (Fig. 8a).To further confirm the ES-induced root growth phenotype, we repeatedly applied 12-h d À1 ES treatment for 4 d and measured the length of the primary roots.This iterative treatment showed the primary root of pE2Fa::E2Fa T314AT315A -HA/e2fa plants grew apparently longer than did the pE2Fa::E2Fa-MYC/ e2fa under ES conditions (Fig. 8b,c).

New Phytologist
To provide further evidence of protein phosphorylationdependent degradation of E2Fa by SnRK1.1, we generated several F 1 double-heterozygous transgenic lines by crossing E2Fa or E2Fa T314AT315A /e2fa with Col-0 or SnRK1.1 WT transgenic plants (Fig. S10).Protein accumulation of E2Fa was compromised by SnRK1.1 in the double-heterozygous transgenic lines under normal growth condition.Notably, protein accumulation of E2Fa T314AT315A was higher than that of E2Fa WT , consistent with the finding that SnRK1.1 phosphorylates T314 and T315 of E2Fa to trigger subsequent degradation of the protein (Fig. 8d).Furthermore, phosphorylation-dependent shift was not shown when SnRK1.1 is co-expressed with E2Fa T314AT315A (Fig. 8e).
Even though E2Fa T314AT315A protein stability was higher than that of E2Fa in the presence of SnRK1.1, the protein stability was not robust in the crossed line (Fig. 8d).To find the reason, we repeated the experiment in the presence of mitogen-activated protein kinase (MPK) inhibitor PD98059 since MPK is activated by SnRK1.1 (Cho et al., 2016) and E2Fa has six putative MPKbinding motifs (ELM: http://elm.eu.org/).With the MPK inhibitor, E2Fa T314AT315A was not degraded in the crossed line with SnRK1.1 WT (Fig. 8f), suggesting that MPK-mediated phosphorylation-dependent degradation of E2Fa may be involved in the weak protein stability.
As expected, the primary roots of SnRK1.1 WT plants, Col-0 9 SnRK1.1 WT plants, and SnRK1.1 WT 9 pE2Fa::E2Fa-MYC/ e2fa plants were shorter than those of WT Col-0 (Fig. 8g), further confirming that SnRK1.1 promotes its protein degradation via phosphorylation of E2Fa at the T314/T315 residues, resulting in S-phase suppression and root growth restriction.In Arabidopsis, plant glucose-metabolic sensor TARGET OF RAPAMYCIN (TOR) kinase phosphorylates and activates E2Fa/ b, which leads to the progression of S-phase in the meristem cells.Activation of TOR-E2F mechanism requires both glucose and light signals at the shoot apex, while glucose energy signal alone is sufficient at the root apex (Li et al., 2017).When the transcription factor E2Fa is phosphorylated by TOR kinase, it promotes G1-S transition, resulting in Arabidopsis root growth promotion (Xiong et al., 2013).Previous studies showed that AMPKmediated REGULATORY-ASSOCIATED PROTEIN OF TOR (RAPTOR) phosphorylation represses the TOR kinase activity and cell cycle under starvation conditions (Gwinn et al., 2008).In plants, SnRK1.1 interacts with RAPTOR1B in vivo and phosphorylated it in vitro (Nukarinen et al., 2016).Therefore, this mechanism was suggested to be conserved in plants despite differences in detail (Gonzalez et al., 2020).Furthermore, SnRK1.1 is translocated from nucleus to cytosol and reduces TOR activity in ABA response (Belda-Palaz on et al., 2022).However, crosstalk mechanisms of SnRK1 and TOR signaling in ES are elusive largely in plants.In the current study, we showed that SnRK1.1 directly phosphorylates the MB domain of E2Fa and subsequently degrades E2Fa protein.Since TOR-E2Fa signaling is a key factor for root stem cell activity, SnRK1-mediated E2Fa degradation results in the inhibition of TOR-E2Fa signaling in an antagonistic manner.Taken together, the signaling protein kinases as fundamental regulatory modules of cell cycle dynamically integrate cellular energy status into the fine-tuning of plant organ growth.
A mammalian SNF1 homolog AMPK serves as a critical cancer repressor to inhibit G1-S transition and cell proliferation (Hardie & Alessi, 2013).Growth inhibition and/or arrest is commonly associated with SnRK1-expressing transgenic plants as well (Tom e et al., 2014;Broeckx et al., 2016;Hulsmans et al., 2016;Muralidhara et al., 2021).Counterintuitively, SnRK1 has also been reported to phosphorylate KRP6 and promote cell cycle progression (Guerinier et al., 2013).Since diverse cellular processes including apoptosis and autophagy are most likely triggered by KPR6 expression (Liang et al., 2007), SnRK1-inducible growth in KRP6-expressing transgenic plants needs to be further tested whether cell cycle progression is the major cause of plant growth restoration and whether SnRK1 could either promote or terminate cell proliferation in different cellular contexts.
Although differential regulation of SnRK1.1 and SnRK1.2 was reported in various stress conditions (Fragoso et al., 2009;S anchez-Villarreal et al., 2018), SnRK1.1 and SnRK1.2 are functionally conserved as an a-subunit of SnRK1 (Baena- Gonzalez et al., 2007).In our study, the protein levels and activities of both SnRK1s were increased by ES conditions.Furthermore, singleknockout mutants (snrk1.1 or snrk1.2) did not produce any obvious phenotype by ES conditions (Fig. S2), reflecting their functional redundancy.For this reason, we relied on the use of SnRK1.1 IN conferring dominant-negative effect on SnRK1s.In addition, interestingly, in the SnRK1.1 WT and SnRK1.1 IN overexpressing transgenics, the endogenous SnRK1.1 proteins were not detected (Fig. 1c).This phenomenon was also shown in two different SnRK1 overexpressing lines (Baena-Gonzalez et al., 2007;Cho et al., 2016).Since SnRK1 is a central regulator of ES signaling, it is prudent to think that there is a negative feedback mechanism controlling their endogenous protein level or activity through promoter activity or mRNA nontranslational regions (Van der Krol et al., 1990;Betti et al., 2021).However, it is remained to be uncovered.
SnRK1 kinase activation/phosphorylation is well documented under various extracellular stress conditions (Baena- Gonzalez et al., 2007;Fragoso et al., 2009;Crozet et al., 2014;Sheen, 2014;Cho et al., 2016;Belda-Palaz on et al., 2020;Gutierrez-Beltran et al., 2021).The protein kinase activity, activation of which is likely linked to its phosphorylation, is clearly involved in Arabidopsis organ growth modulation under ES conditions, at least in the early seedling stage (4-10 d after germination).Further mechanistic understanding of protein complex of SnRK1.1 is required to evaluate whether SnRK1.1 phosphorylation is attributed to its kinase activation.In the current study, SnRK1.1 kinase activity involved in plant growth restriction has been genetically validated with hyposensitive growth responses of SnRK1.1 IN transgenic plants.Protein turnover and organ growth analyses of F 1 double-heterozygous transgenic plants expressed with SnRK1.1 WT and E2Fa or E2Fa T314AT315A have further provided compelling evidence for SnRK1.1 kinase-inducible phosphorylation-coupled E2Fa degradation, resulting in plant growth restriction.Identification of the E3 ligase(s) involved in the ubiquitination of phosphorylated E2Fa will yield insights into the detailed mechanism for protein degradation.
Likewise, E2Fb protein stability and its transcriptional activity were negatively regulated by SnRK1.1 (Figs 3, 5).However, its spatial gene expression profile was not correlated with that of SnRK1.1 in root (Fig. S4d).While both E2Fa and E2Fb are involved in the coordination of cell proliferation, it was suggested that E2Fa and E2Fb could have different functions based on their association with components of the evolutionary conserved multi-subunit DP-Rb-E2F and MuvB complex (Kobayashi et al., 2015;Leviczky et al., 2019).Therefore, our finding that SnRK1.1 differentially regulates E2Fa and E2Fb may support the hypotheses of differential function of the two.
Phosphorylation-dependent conformational change affects the transcriptional activity or binding affinity of transcription factors (Kedage et al., 2017;Mizoi et al., 2019).MB domain of E2Fa is important for interaction with transcriptional partners (Black et al., 2005).In the current study, we show that SnRK1.1 phosphorylated the MB domain of E2Fa (Fig. 4), which resulted in the degradation of E2Fa (Fig. 5).It is prudent to think that the SnRK1.1-mediatedphosphorylation of the MB domain of E2Fa might affect its binding affinity with the transcriptional partner (s).However, the phosphorylation did not affect the transcriptional activity of E2Fa, evidenced by the high level of LUC reporter gene activity with the co-expression of SnRK1.1 in the presence of MG132 (Fig. 5c).This suggests that the SnRK1.1mediatedphosphorylation may not have an impact on the binding affinity of E2Fa with its transcriptional partner protein(s), while it leads to proteasome-dependent degradation of the protein.
In conclusion, our study has unraveled a novel and direct regulation of E2F transcription factors by SnRK1.1, which allows the plants to dynamically adjust their vegetative growth in response to cellular energy status, hence serving as a robust adaptation mechanism to cope with constantly fluctuating environmental conditions.

Fig. 1
Fig. 1 Cellular energy stress (ES) induces organ growth restriction in Arabidopsis.(a, b) Protein blot (a) and quantitative (b) analysis of total SnRK1.1 and phosphorylated SnRKs (P-SnRK1.1 and P-SnRK1.2) in Col-0 after ES treatment for 24 h on 4-d-old seedlings grown in ½MS liquid medium containing 0.5% sucrose.Total protein from the plants was used in the protein blot analysis with anti-pT172-AMPKa antibody for the detection of phosphorylated SnRKs and anti-SnRK1.1 for protein level of SnRK1.1.Actin11 was used as a control.Values are mean ± SD.Asterisks indicate values statistically different from controls (t-test; **, P < 0.01; *, P < 0.05).(c) Protein blot analysis of SnRK1.1 and phosphorylated SnRK1.1 (P-SnRK1.1) of 5-d-old Col-0, transgenic plants expressing SnRK1.1 WT and SnRK1.1 IN .Total protein from the plants was used in the protein blot analysis with anti-pT172-AMPKa antibody and anti-HA antibody.Actin11 was used as a control.(d) Growth phenotypes of 3-wk-old Col-0 and SnRK1.1 WT -, or SnRK1.1 IN -expressing transgenic plants grown on soil.Bar, 20 mm.(e) Primary root growth and quantitative analysis of Col-0 plants expressing SnRK1.1 WT and SnRK1.1 IN .The root length of 5-dold plants grown in ½MS liquid medium containing 0.5% sucrose was measured.Values are mean ± SD.Different letters indicate statistical differences according to ANOVA (P < 0.05).All experiments were repeated at least three times with similar results.(f) Cellular images of EdU staining of the root apical meristems of 5-d-old Col-0 and SnRK1.1 WT or SnRK1.1 IN -expressing transgenic plants grown in ½MS liquid medium containing 0.5% sucrose.Bright-field images served as a control.Arrowheads indicate quiescent center.BF, bright field; Bar, 100 lm.(g) Expression analysis of cell cycle-related genes of 5-d-old Col-0 and SnRK1.1 WT or SnRK1.1 IN -expressing transgenic plants using RT-qPCR.The plants were grown in ½MS liquid medium containing 0.5% sucrose for 5 d, and total RNA was isolated from the plants and synthesized cDNA with the RNA.EIF4a was used as a control.Values are mean ± SD.Asterisks indicate values statistically different from Col-0 control (t-test; *, P < 0.05).
SnRK1.1 IN and pE2Fa::gE2Fa-GFP transgenic plants (i.e.genetic complementation of e2fa) were larger and longer, respectively, than those of the SnRK1.1 WT plants.In light of the fact that SnRK1.1 negatively regulates both organ development and ESinduced senescence and inflorescence development (Baena-

Fig. 5
Fig. 5 SnRK1.1 degrades E2Fa/b proteins.(a) Protein blot analysis of E2Fa and E2Fb with a combination of SnRK1.1 WT and SnRK1.1 IN co-expression using anti-MYC antibody.C-terminal MYC-conjugated E2Fa and E2Fb were co-expressed in the protoplasts with SnRK1.1 WT or SnRK1.1 IN for protein blot analysis using anti-MYC antibody.Actin11 served as a control.(b) Cellular images of E2Fa-GFP and E2Fb-GFP with and without SnRK1.1 WT co-expression.Cterminal GFP-conjugated E2Fa and E2Fb wereexpressed in the protoplasts with or without SnRK1.1.Bar, 10 lm.(c) Measurements of E2Fa/b translation reporter activity in the protoplasts with and without SnRK1.1 WT co-expression in the presence and absence of the proteasome inhibitor MG132.C-terminal fLUC-conjugated E2Fa/b were expressed in the protoplasts under HBT promoter with or without SnRK1.1 and MG132.pUBQ10-rLUC served as a control.Values are mean ± SD.Asterisks indicate values statistically different from control (t-test; **, P < 0.01).(d) Protein blot analysis of E2Fa-MYC and E2Fb-MYC with or without SnRK1.1 WT and RBR-HA co-expression using anti-MYC antibody.Actin11 served as a control.

Fig. 7
Fig. 7 SnRK1.1-dependentphosphorylation of E2Fa residues T314 and T315 leads to E2Fa protein degradation.(a) Promoter activities of ETG1 with combination of E2Fa, E2Fa variants, and SnRK1.1 co-expression in protoplasts.pETG1-fLUC was transfected to protoplasts with designated effector combinations.pUBQ10-rLUC served as a control.Values are mean ± SD.Asterisks indicate values statistically different from control (t-test; **, P < 0.01).(b) Protein blot analysis for protein accumulation of E2Fa and E2Fa variants with or without SnRK1.1 coexpression in protoplasts.Protein blot analysis was carried out using anti-HA antibody.Actin11 served as a control.(c) Protein blot and quantitative analyses of cell-free protein degradation assay of GST-E2Fa with the protein extracts of Col-0 or SnRK1.1 WT transgenic plants in the presence and absence of MG132.Actin11 served as a control.Values are mean AE SD.(d) Protein blot and quantitative analyses of cell-free protein degradation assay of GST-E2Fa and GST-E2Fa T314AT315A with the protein extracts of SnRK1.1 WT transgenic plants in the absence and presence of MG132.Actin 11 served as a control.Values are mean ± SD.Asterisk indicates values statistically different (t-test; *, P < 0.05).(e) Protein blot and quantitative analyses of E2Fa in pE2Fa::E2Fa-MYC or pE2Fa::E2Fa T314AT315A -HA transgenic plants after energy stress (ES) treatment.Four-day-old seedlings were treated with ES for indicated times.Total protein was isolated from the plants and carried out protein blot analysis with anti-MYC antibody.Actin11 served as a control.Values are mean ± SD.Asterisks indicate values statistically different (t-test; *, P < 0.05).(f) Gene expression analysis of ETG1 and MCM5 in pE2Fa::E2Fa-MYC or pE2Fa:: E2Fa T314AT315A -HA-expressing plants after ES treatment.Four-day-old seedlings were treated with ES for 12 h and carried out RT-qPCR.EIF4a served as a control.Values are mean ± SD.Asterisks indicate values statistically different from Col-0 control (ttest; *, P < 0.05).[Correction added after online publication 7 December 2022: labels in panels (b, e, f) have been updated.] New Phytologist (2023) 237: 823-839 www.newphytologist.comÓ 2022 The Authors New Phytologist Ó 2022 New Phytologist Foundation.