PI4Kγ2 interacts with E3 ligase MIEL1 to regulate auxin metabolism and root development.

Root development is important for normal plant growth and nutrient absorption. Studies have revealed the involvement of various factors in this complex process, improving our understanding of the relevant regulatory mechanisms. Here, we functionally characterize the role of Arabidopsis phosphatidylinositol 4-kinase γ2 (PI4Kγ2) in root elongation regulation, which functions to modulate stability of the RING-type E3 ligase MYB30-INTERACTING E3 LIGASE 1 (MIEL1) and auxin metabolism. Mutant plants deficient in PI4Kγ2 (pi4kγ2) exhibited a shortened root length and elongation zone due to reduced auxin level. PI4Kγ2 was shown to interact with MIEL1, regulating its degradation and further the stability of transcription factor MYB30, which suppresses auxin metabolism by directly binding to promoter regions of GH3.2 and GH3.6. Interestingly, pi4kγ2 plants presented altered hypersensitive response, indicating that PI4Kγ2 regulates synergetic growth and defense of plants through modulating auxin metabolism. These results reveal the importance of protein interaction in regulating ubiquitin-mediated protein degradation in eukaryotic cells, and illustrate a mechanism coordinating plant growth and biotic stress response.


Introduction
The phosphatidylinositol (PI) signaling pathway is a vital regulator of biological function in eukaryotes.Phosphoinositides have long been recognized as major components of eukaryotic membranes essential for many functions in eukaryotes, including cell signaling, membrane dynamics, and trafficking (Ischebeck et al., 2010).The functions of the PI signaling pathway are less understood in plants compared with mammals and yeast.
Although physiological roles of type-III PI4Ks have been reported, most of the eight type-II PI4Ks in Arabidopsis have not been functionally characterized.Type-II PI4Ks contain PI3/4 kinase domains and variable numbers (none, one, or two) of ubiquitin-like (UBL) domains, whereas they lack the PI-binding domains such as PH (in PI4Kα) or PPC domains (in PI4Kβ, Mueller-Roeber, 2002).Lipid kinase activity has not previously been detected in type-II PI4Ks, whereas Arabidopsis PI4Kγ3, PI4Kγ4, and PI4Kγ7 were shown to possess autophosphorylation activity.AtPI4Kγ4 phosphorylates serine/threonine residues of UFD1 (ubiquitin fusion degradation 1) and RPN10 (regulatory particle non-ATPase 10) (Galvão et al., 2008).TaPI4KIIγ has been confirmed as a stress-inducible gene in wheat (Triticum aestivum), which interacts with and phosphorylates wheat ubiquitin fusion degradation protein (TaUDF1) (Liu et al., 2013); however, the effects of UFD1 and RPN10 phosphorylation remain unknown.AtPI4Kγ3 selectively binds to a few PIs and is important for reinforcing plant responses to environmental stresses and in the delay of floral transition (Akhter et al., 2016).AtPI4Kγ5 interacts with the membrane-bound transcription factor ANAC078 to regulate auxin biosynthesis and leaf margin development (Tang et al., 2016).
These studies indicate that type-II PI4Ks participate in the regulation of specific physiological processes through different mechanisms.
The functionality of proteins is tightly regulated by various kinds of post-translational modifications.Ubiquitination is an essential regulatory mechanism that alters protein stability by mediating the proteasomal degradation of target proteins (Schwechheimer et al., 2018).Ubiquitination involves E1 ubiquitin activating enzymes, E2 ubiquitin conjugating enzymes, and E3 ligases (Rape et al., 2018).E3 ligases recognize and bind specifically to the substrate (Vierstra et al., 2009), and thus play key roles in protein degradation.Although an increasing number of reports show the crucial roles of E3 ligases in plants, only a few reports have focused on the regulation of E3 ligases at the post-translational level.Indeed, in addition to transcriptional regulation, E3 ligases are subjected to various post-translational modifications.Tyrosine phosphorylation of the E3 ligase GARU at Tyr321 inhibits the GARU-GID1A interaction and hence degradation of GID1 (Nemoto et al., 2017).CSN (Cop9 Signalosome), which mediates the deneddylation of cullins, is required to inactivate CRL (Cullin-RING E3 ubiquitin-Ligases) complexes, and then deter substrate recognition and prevent CRL subunit instability by counteracting their auto-ubiquitination and subsequent degradation (Merlet et al., 2009).Pirh2 protein is a p53-induced E3 ligase regulated by different proteins: CaMK II and Cdk9 phosphorylate Pirh2 then abrogate its E3 ligase activity toward p53 (Duan et al., 2007;Bagashev et al., 2013); TIP60 (Tat-interactive protein of 60 kDa) interacts with Pirh2 to alter the subcellular localization of Pirh2 (Logan et al., 2004); and PLAGL2 interacts with Pirh2 to prevent the proteasomal degradation of Pirh2 (Zheng et al., 2007).These findings indicate that post-translational modification is critical in E3 functions.
Many signaling pathways control root growth and development, particularly the phytohormone auxin, which regulates both cell proliferation and elongation of root cells (Street et al., 2016).Previous studies showed that AUX1 mediates the inhibitory effects of cytokinin on root cell elongation (Street et al., 2016;Péret et al., 2012).YDK1, a GH3 gene, might function as a negative regulator of root cell elongation by controlling the level of the active form of auxin (Takase et al., 2004).In rice (Oryza sativa), OsARF12, a transcription activator of auxin-response genes, mediates root cell elongation by inhibiting most OsYUCCAs auxin-synthesis genes (Qi et al., 2012).In addition, phospholipase D-derived PA (phosphatidic acid) promotes root hair development under phosphorus deficiency by suppressing vacuolar degradation of PIN2 (Lin et al., 2020).PIP5K2 (phosphatidylinositol monophosphate 5-kinase 2) and PIP5K1 are involved in root development regulation through regulation of PIN proteins (Mei et al., 2012) and PIN-mediated auxin efflux (Barbosa et al., 2016).
Recent genetic studies have revealed important roles of distinct E3 ligases in phytohormone signaling pathways (Kelley et al., 2018), operating via the degradation of various components, including receptors, positive or negative regulators, or those related to hormone metabolisms.Through systematic biochemical and genetic studies, we here report the functional characterization of an Arabidopsis type-II PI4K, PI4Kγ2, that regulates root growth through its interaction with the ubiquitin E3 ligase MYB30-INTERACTING E3 LIGASE 1 (MIEL1).MIEL1 is a negative regulator of hypersensitive response (HR) cell death by promoting MYB30 turnover in Arabidopsis (Marino et al., 2013).MIEL1 protein level is significantly reduced in pi4kγ2 mutant plants and MYB30 directly binds to promoter regions of GH3.2 and GH3.6, thereby suppressing auxin metabolism and hence root growth.
Our data demonstrate that PI4Kγ2 is critical for MIEL1 stability and MYB30 turnover and elucidate how MIEL1 coordinates auxin-dependent plant growth and defensive HR in Arabidopsis.

Deficiency of PI4Kγ2 results in short roots
Type-II PI4Ks, which lack the PI-binding domains, play roles in plant development via protein-protein interaction or protein phosphorylation (Mueller-Roeber, 2002;Galvão et al., 2008;Liu et al., 2013).Our previous study showed that Arabidopsis PI4Kγ5 regulates auxin biosynthesis and leaf margin development by interacting with and regulating the cleavage of the membrane-bound transcription factor ANAC078 (Tang et al., 2016).To expand our understanding of type-II PI4K functions, we further functionally characterized AtPI4Kγ2.
Histochemical analysis of transgenic lines harboring a PI4Kγ2 promoter-reporter construct (Glucuronidase, pPI4Kγ2::GUS) showed that PI4Kγ2 is expressed in young seedlings, floral tissue, rosette leaf, cauline leaf, stem, and inflorescence (Fig. 1B).GUS signals were predominantly observed in the root apical region of young seedlings (Fig. 1B, 1), suggesting a potential role of PI4Kγ2 in root development.
A T-DNA insertional mutant pi4kγ2 (Alonso et al., 2003) was identified, which we used to investigate the physiological role of PI4Kγ2.T-DNA was inserted in the first exon of PI4Kγ2 (Fig. S1A, B) and RT-qPCR analysis confirmed the deficient transcription of PI4Kγ2 in homozygous mutant plant (Fig. S1C), indicating that pi4kγ2 is a knockout mutant.
Phenotypic observation showed that growth of the pi4kγ2 mutant is similar to wild type; however, interestingly, it presents short roots (Fig. 1C).Considering that primary root growth is dependent on both cell proliferation and cell elongation, detailed analysis revealed a decreased length of root elongation zone and decreased root epidermal cell length in the maturation zone for pi4kγ2 (Fig. 1D).Considering that there was no change of length of meristematic zone of pi4kγ2 roots (Fig. S1D), these results suggest that PI4Kγ2 regulates root growth by modulating cell length.As expected, complementation studies demonstrated that reinstated PI4Kγ2 expression (Fig. S2) resulted in recovered root growth, confirming the role of PI4Kγ2 in root development.

PI4Kγ2 interacts with MIEL1
To investigate the functional mechanism of how PI4Kγ2 regulates root cell elongation, yeast two-hybrid screening was performed using whole PI4Kγ2 protein as bait to search for PI4Kγ2-interacting partners.Interestingly, MIEL1, a zinc finger RING-type protein localized to the cytoplasm and nucleus (Marino et al., 2013), was identified (Fig 2A).Subcellular localization studies by observing fluorescence showed the distribution of YFP (yellow fluorescent protein)-PI4Kγ2 fusion protein in cytoplasm and nucleus of Arabidopsis cells, which is similar as that of MIEL1 (Fig. 2B, upper).Transient expression of YFP-PI4Kγ2 and CFP-MIEL1 fusion proteins in Arabidopsis protoplasts confirmed that PI4Kγ2 co-localizes with MIEL1 (Fig. 2B, bottom), further suggesting a possible interaction between them, which was investigated by using BiFC (bimolecular fluorescence complementation) assays.Indeed, strong fluorescence was detected in the nucleus and cytoplasm of cells co-expressing cYFP-MIEL1 and nYFP-PI4Kγ2 (Fig. 2C), confirming the PI4Kγ2-MIEL1 interaction in vivo.

Deficiency of MIEL1 and MYB30 results in shorter or longer roots, respectively
Previous studies showed that MIEL1 interacts with and ubiquitinates MYB30, leading to MYB30 proteasomal degradation and suppressed plant defense responses (Marino et al., 2013).Considering that PI4Kγ2 interacts with MIEL1, we thus investigated the function of MIEL1 and MYB30 in root growth.Observation of growth of miel1 (Marino et al., 2013) and myb30 knockout lines (Liu et al., 2014) showed that, similar to pi4kγ2 mutant, miel1 exhibited shortened roots and decreased length of root elongation zone and epidermal cells of the root maturation zone, whereas myb30 showed longer root length and increased length of root elongation zone and epidermal cells of root maturation zone (Fig. 3), which is consistent with MIEL1-facilitated MYB30 degradation and indicates a specific role for MIEL1 and MYB30 in regulating root growth.Genetic analysis showed that miel1 pi4kγ2 double mutants displayed similar root growth as miel1 and pi4kγ2, whereas pi4kγ2 myb30 or miel1 myb30 double mutants presented root growth similar to myb30 (longer roots), indicating the genetic epistasis of MYB30 and that root growth regulation by PI4Kγ2 and MIEL1 is achieved via the effect of MIEL1 on MYB30.

PI4Kγ2 regulates MIEL1 stability and accelerates MYB30 turnover in planta
Many reports in animals have shown that functions of E3 ubiquitin ligase can be regulated at the post-translational level, including deneddylation, phosphorylation, and interaction with other proteins.To test whether MIEL1 is regulated by PI4Kγ2 through direct interaction in planta, MIEL1 fused with mCherry was expressed in WT and pi4kγ2 to study the effects of PI4Kγ2 on MIEL1 in vivo.
A quick cross was performed to obtain the isogenic MIEL1-mCherry plants.Pollens from MIEL1-mCherry (in pi4kγ2) homozygous lines were respectively spread on the stigmas of pi4kγ2 and WT, and analysis of F1 plants with comparable MIEL1 expression levels by immunoblot analysis showed that MIEL1-mCherry accumulated less in pi4kγ2 compared to that in PI4Kγ2/pi4kγ2 (Fig. 4A), suggesting that PI4Kγ2 functions in regulating MIEL1 stability.Consistently, transient expression of FLAG-tagged MIEL1 alone or with Myc-tagged PI4Kγ2 in Nicotiana benthamiana leaves further showed greater accumulation of MIEL1 protein following the co-expression of PI4Kγ2 (Fig. S3A).These results indicate that PI4Kγ2 positively regulates MIEL1 stability through direct interaction.
MIEL1 mediates the proteasomal degradation of MYB30 (Marino et al., 2013).We thus examined whether PI4Kγ2-MIEL1 interaction influences the degradation and turnover of MYB30.Analysis of N. benthamiana leaves transiently expressing FLAG-tagged MYB30 showed that MYB30 protein accumulated after treatment with proteasome inhibitor MG132 and was reduced following the co-expression of MIEL1 (Fig. 4C), which is consistent with the previous reports (Marino et al., 2013).Interestingly, MYB30 protein level was significantly reduced when co-expressed with PI4Kγ2 and MIEL1 (Fig. 4C), indicating that PI4Kγ2 regulates the turnover of MYB30 through interacting with and stabilizing MIEL1.

PI4Kγ2 regulates auxin metabolism through MYB30
Root growth is a complex process and regulated by various developmental signals and environmental factors, particularly phytohormones.Auxin level, signaling, and polar transport are involved in root growth regulation.Considering that free IAA plays main roles in auxin effects, we first investigated whether IAA level is altered in pi4kγ2 by liquid chromatography-mass spectrometry (LC-MS) analysis and results revealed decreased free IAA content in pi4kγ2 roots (Fig. 5A).Consistently, the short-root phenotype of pi4kγ2 and miel1 was recused by exogenous IAA and both mutants presented similar root length under 0.01 μM IAA (Fig. 5B), confirming that reduced IAA level leads to the short-root phenotype of pi4kγ2 and miel1.
Considering the genetic epistasis of MYB30 on PI4Kγ2 and MILE1, and that MIEL1 mediates the proteasomal degradation of MYB30, it was hypothesized that MYB30 negatively regulates auxin metabolism and hence cell elongation.The chemical compound L-Kynurenine (L-Kyn, an inhibitor of TAA1/TAR to suppress auxin synthesis, He et al., 2011) was utilized to test whether the phenotype of myb30 was the result of increased auxin accumulation.Indeed, the longer root of myb30 was suppressed under L-Kyn treatment (similar root length was observed at 2 μM L-Kyn, Fig. 5C), suggesting the longer roots of myb30 might be due to increased auxin content.This was confirmed by increased free IAA level of myb30 roots demonstrated by LC-MS (Fig. 5D).As expected, RT-qPCR analysis of expression of some IAA metabolism-related genes showed that GH3 members (GH3.1,GH3,2,GH3.4,GH3.6,GH3.7,GH3.13,GH3.17,GH3.18,and GH3.19, encoding auxin-metabolism enzymes) presented decreased expression levels, whereas YUC4 and YUC7 (encoding key enzymes in auxin biosynthesis) presented increased expression levels in myb30 roots (Fig. 5E).Further analysis showed that expression levels of GH3 members (GH3.1,GH3.2, GH3.4,GH3.17,GH3.18,and GH3.19) were increased in pi4kγ2 and miel1 roots (Fig. 5F).GH3 proteins synthesize IAA-amino acid conjugates, either the intermediate destined for IAA metabolism or the inactive storage form of IAA, to reduce the free IAA level.Increased (or decreased) expression levels of GH3s are consistent with the reduced auxin contents in pi4kγ2 and miel1 (or myb30) and confirmed that PI4Kγ2 regulates auxin metabolism and root growth by interacting with MILE1 to modulate MYB30.
MYB30 is a R2R3-MYB transcriptional activation factor involved in the very first step of HR (Vailleau et al., 2002).Interestingly, prediction analysis (AGRIS, http://arabidopsis.med.ohio-state.edu/)revealed the presence of putative binding sequences of MYB members (Supplemental Table S1).A ChIP-qPCR analysis with an anti-Flag antibody was subsequently performed to investigate whether MYB30 directly binds the promoter of these genes.A MYB30-Flag fusion protein (driven by CaMV35S promoter) was expressed in WT seedlings and two transgenic lines with comparable protein expression levels were selected for a ChIP assay.qPCR analysis revealed that distinct DNA fragments of GH3.2 (P2, Fig. 5G) and GH3.6 (P3, Fig. 5G) promoters were more enriched, indicating that MYB30 regulates the expression levels of GH3.2 and GH3.6 genes by directly binding their promoters.

PI4Kγ2 negatively regulates resistance and hypersensitive response following bacterial inoculation
MIEL1 attenuates HR to bacterial infection, consistent with the enhanced or reduced responses of myb30 or MYB30-overexpression lines (Marino et al., 2013).Observation showed that compared to chlorosis symptoms of WT with Pseudomonas syringae pv.tomato AvrRpm1 (Pst AvRpm1), miel1 plants showed clear HR cell death symptoms (Fig. 6A).
Similar to miel1, pi4kγ2 presented HR cell death symptoms as well at 64 hours after infection (hpi) with Pst AvrRpm1 (Fig. 6A), indicating the important function of PI4Kγ2 in plant response to bacterial inoculation.Consistent with the faster HR, pi4kγ2 showed increased resistance in response to inoculation with Pst AvrRpm1 compared to WT plants (Fig. 6B), confirming that PI4Kγ2 functions as a negative regulator of plant defense.
RT-qPCR analysis of expression of MYB30 VLCFA (very long chain fatty acid)-related target genes, specifically KCS1 and FDH (Marino et al., 2013), showed their increased expression in pi4kγ2 and miel1 plants compared to WT plants following bacterial inoculation (Fig. 6C), further indicating that PI4Kγ2 negatively mediates Arabidopsis defense response, possibly by regulating MYB30 stability.

Discussion
Large numbers of E3 ubiquitin ligases (~1,500 E3-encoding genes are predicted in the Arabidopsis genome, Guzman, 2014) regulate multiple cellular and biological processes by interacting with and degrading distinct target proteins (Merlet et al., 2009).Many reports have shown that E3 ligases can be regulated through post-translational modifications in animals, which remains relatively undocumented in plants.Our studies functionally characterized the role of PI4Kγ2, an Arabidopsis atypical type-II PI4K, in root elongation and indicated that PI4Kγ2 is an E3 ligase mediator that inhibits ubiquitination of E3 (MIEL1) through protein-protein interaction.These findings provide further insights into Arabidopsis type-II PI4Ks and help to elucidate the regulatory mechanism of plant growth/development control, particularly the balance between plant resistance and development, at a post-translational level (Fig. 6D).
MIEL1 is a C3H2C3 canonical RING-type E3 ligase and plays important roles in ABA signaling and plant defense responses; however, the upstream regulation of MIEL1 remains poorly understood.A previous study showed that MIEL1 can be ubiquitinated (Lee & Seo, 2016).We found that PI4Kγ2 ensures the reduced ubiquitination and enhanced stability of MIEL1 (Fig. 4) through their interaction, leading to stimulated MYB30 degradation, providing additional evidence as to how type-II PI4Ks confer their functions in plants.
Similarly, in mammalian cells, interaction between PLAGL2 and Pirh2 (Pirh2 encodes a RING-H2 domain-containing protein with intrinsic ubiquitin-protein ligase activity) prevents the proteasomal degradation of Pirh2 (Zheng et al., 2007).These results suggest that protein-protein interaction is a general regulatory mechanism for ubiquitination and stability of E3 ligases; however, the exact function of ubiquitin-like (UBL) domains of type-II PI4Ks and the detailed mechanisms behind this regulation require further investigation.
MYB30 is a pleiotropic mediator that regulates a variety of physiological processes and signaling responses, including pathogen-induced HR, flowering time, and brassinosteroid and abscisic acid signaling (Raffaele et al., 2006;Raffaele et al., 2008;Li et al., 2009;Zheng et al., 2012;Raffaele & Rivas, 2013;Liu et al., 2014).However, its function in root cell elongation is unclear.Our studies show that MYB30 deficiency results in the promoted cell elongation of roots by altering auxin metabolism, which is consistent with the decreased expression of many GH3s and reveals a further function of MYB30 in regulating auxin metabolism.
Understanding control of the balance between plant resistance and development is fundamental and it was recently shown that GA-ABA and auxin-JA synergistically regulate resistance and growth (Yuan et al., 2018).Interestingly, we found that PI4Kγ2-MIEL1-MYB30 regulates auxin-mediated growth and HR, and MYB30 may act in crosstalk of growth and defense signaling.MYB30 positively regulates HR cell death in response to pathogen attack (Vailleau et al., 2002) while inhibiting growth by decreasing auxin level.Indeed, auxin is an important plant growth regulator and also influences plant-pathogen interactions, and previous studies showed that elevated IAA biosynthesis in plants led to enhanced susceptibility to DC3000 (Mutka et al., 2013).Considering its myriad of regulatory targets and differential expression of PI4Kγ2-MIEL1-MYB30 after inoculation with Pst AvrRpm1, i.e.PI4Kγ2 and MYB30 expression was specifically induced at 4 hpi and 1 hpi, respectively, whereas MIEL1 expression was repressed after inoculation at 1 hpi (Fig. S4), it would seem that MYB30 acts on upstream effectors of auxin signaling and HR progress.
In sum, PI4Kγ2 interacts with MIEL1, thus inhibiting its turnover and leading to increased proteasomal degradation of MYB30, which attenuates auxin metabolism and hence increases auxin level to avoid superfluous activation of HR.Under PI4Kγ2 deficiency, MIEL1 is ubiquitinated and degraded, thereby resulting in accumulated MYB30 protein and hence activated HR and decreased auxin level/suppressed growth (Fig. 6D).Increased auxin enhances plant growth and auxin-mediated susceptibility, whereas decreased auxin results in growth inhibition and enhanced disease resistance.Considering the positive effect of PI4Kγ2 on MIEL1 stability and negative effect of MIEL1 on MYB30 activity, these results manifest that PI4Kγ2 may contribute to synergistic regulation of plant defense and growth.
Seeds were surface-sterilized and sown on plates containing Murashige and Skoog (MS) medium (Duchefe Biochemie, The Netherlands).After stratification at 4°C for 3 days, seedlings were grown in phytotron with a 16-h light/8-h dark cycle (22°C) for normal growth and seed harvesting.

Identification of T-DNA mutants
Mutant pi4kγ2 carrying a T-DNA insertion in the first exon was confirmed by PCR amplification using primers PI4Kγ2-L and PI4Kγ2-R.Transcription level of PI4Kγ2 was examined by RT-qPCR using primers PI4Kγ2-3 and PI4Kγ2-4.T-DNA insertion of mutant miel1 was confirmed by PCR amplification using primers MIEL1-1 and MIEL1-2.T-DNA insertion of mutant myb30 was confirmed by PCR amplification using primers MYB30-1 and MYB30-2.All primers used in this study are listed in Supplemental Table S2.

Constructs and plant transformation
For expression of PI4Kγ2 in WT or pi4kγ2, coding sequence of PI4Kγ2 (1-903 bp) was amplified with primers (PI4Kγ2-5 and PI4Kγ2-6) and subcloned into the mCherry vector with C-terminal fusion.For expression of cMyc-PI4Kγ2 in Nicotiana benthamiana, the coding sequence of PI4Kγ2 was amplified (primers PI4Kγ2-7 and PI4Kγ2-8) and subcloned into a modified pEGAD-4XcMyc vector with N-terminal fusion.For expression of MIEL1-Flag in N. benthamiana, the coding sequence of MIEL1 was amplified by PCR (primers MIEL1-3 and MIEL1-4) and subcloned into 1306-Flag vector with C-terminal fusion.For expression of MIEL1-mCherry in WT, pi4kγ2, or N. benthamiana, the coding sequence of MIEL1 was amplified by PCR (primers MIEL1-5 and MIEL1-6) and subcloned into mCherry vector with C-terminal fusion.For expression of MYB30-Flag in N. benthamiana, the coding sequence of MYB30 was amplified by PCR (primers MYB30-3 and MYB30-4) and subcloned into 1306-Flag vector with C-terminal fusion.

Promoter-reporter gene fusion studies
The promoter of PI4Kγ2 (-2,491 bp upstream of ATG) was amplified (primers PI4Kγ2-P1/PI4Kγ2-P2), then cloned into a modified pCAMBIA1300 vector including a GUS reporter.A histochemical assay of GUS activities was performed according to a previous description (Tang et al., 2016).Arabidopsis transformation samples were observed using DIC microscopy (Nikon SMZ1500).

Measurement of free IAA contents by liquid chromatography/ mass spectrometry (LC/MS)
Around 200 mg of roots from 10-day-old seedlings was frozen in liquid nitrogen and ground to a fine powder for free IAA content measurement using a Thermo TSQ Quantum Ultra LC-MS-MS system according to the previous description (Tang et al., 2016).

Immunoblot analysis
Isogenic MIEL1-mCherry plants were first prepared by crossing MIEL1-mCherry (in pi4kγ2) homozygous lines (as the male parent) with WT and pi4kγ2, respectively.Protein extracted from transgenic plants was re-suspended in extraction buffer pH 7.5,150 mM NaCl,, 1 mM EDTA, 1 mM DTT] containing a protease inhibitor cocktail (Roche, Germany).After addition of an equal volume of 2X SDS buffer, the samples were boiled for 5 min, fractionated by 10% SDS-PAGE and transferred to a PVDF membrane by semi-dry blotting.The blots were incubated with a mouse anti-mCherry antibody (Abcam, USA) and then with a bovine anti-mouse IgG AP-conjugated secondary antibody (Santa Cruz Biotechnology, USA).AP activity was detected by BCIP/NBT Detection Reagents (Invitrogen, USA).
For mCherry immunoprecipitation assays, the deubiquitination inhibitor PR-619 (Sigma -Aldrich) was added into the extraction buffer.Supernatants were pre-cleared for 1 h at 4°C with Dynabeads (Life Technology).Then samples were incubated with Anti-mCherry Dynabeads (Life Technology) overnight.All steps were performed at 4°C.Pulled down proteins were eluted by 2X SDS buffer, then boiled for 5 min.MIEL1-immunoprecipitated fractions and ubiquitination of MIEL1 were analyzed by using anti-mCherry or anti-Ub antibody respectively.

BiFC (bimolecular fluorescence complementation) assays
Coding regions of PI4Kγ2 (primers PI4Kγ2-11 and PI4Kγ2-12) and MIEL1 (primers MIEL1-11 and MIEL1-12) were fused with nYFP or cYFP, respectively.Resultant constructs PI4Kγ2-nYFP and MIEL1-cYFP were transformed into Agrobacterium and observed after 48 h according to a previous description (Fang & Spector, 2010).Considering YFP was excited at 488 nm and emission at 510-580 nm, and chlorophyll autofluorescence was emission at 644-714 nm (Body et al., 2019;Vermeer et al., 2006), BiFC results were imaged with an Olympus FV1000 confocal microscope using the 488 nm argon laser and a dichroic filter to visualize YFP.The laser power was set at 5% for the 488 lasers.All images were collected using a 20 x aperture objective.Collection wavelength set at 510-550 nm.

Co-localization studies of PI4Kγ2 and MIEL1
Coding regions of PI4Kγ2 (primers PI4Kγ2-13 and PI4Kγ2-14) and MIEL1 (primers MIEL1-13 and MIEL1-14) were amplified and subcloned into vector pA7 (N-terminus fusion with YFP or CFP respectively), resulting in the YFP-PI4Kγ2 and CFP-MIEL1 fusion constructs, which were transiently expressed in Arabidopsis protoplasts by PEG/CaCl 2 methods (Yoo et al., 2007).The fluorescence was observed by confocal laser scanning microscopy (Olympus FV1000).For YFP/CFP, we used excitation/emission combinations of 514 nm / 530-580 nm for YFP and 458 nm / 470-500 nm for CFP.The laser power was set at 20% for the 458 laser and 5% for the 514 lasers.All images were collected using a 20 x aperture objective.
The assay of kinase activity was performed according to previous description (Tan et al., 2013) with a few modifications.

Bacterial materials
Arabidopsis 4-week-old plants were kept at high humidity 12 h before inoculation and injected with a bacterial suspension of Pst AvrRpm1 at the indicated bacterial densities using a blunt syringe on the abaxial side of the leaves.For measure of in planta bacterial growth, injected leaves samples were harvested 0 and 3 days.A predetermined dilution for each sample was plated on King's B medium and incubated at 28°C for 2 days.Data were submitted to a statistical analysis (Marino et al., 2013).
presented as means ± SD (n=5).Statistical analysis was performed using a two-tailed Student's t-test (*, p < 0.05, compared to WT).B. WT, miel1, and pi4kγ2 seedlings were grown on MS medium supplemented with IAA for 15 days, and root lengths were measured.Experiments were repeated three times and data are presented as means ± SD (n > 30).Statistical analysis revealed significant differences (*, p < 0.05; **, p < 0.01, compared to WT).
C. WT and myb30 seedlings were grown on MS medium supplemented with IAA biosynthesis inhibitor L-Kynurenine (L-Kyn) for 10 days, and root lengths were measured.Experiments were repeated three times and data are presented as means ± SD (n > 30).Statistical analysis revealed the significant differences (*, p < 0.05; **, p < 0.01, compared to WT).D. Quantification of free IAA content by LC-MS revealed the increased IAA level in myb30.Roots of 5-day-old WT and myb30 seedlings were used for analysis and data are presented as means ± SD (n=5).Statistical analysis was performed using a two-tailed Student's t-test (*, p < 0.05, compared to WT).E. Relative expression levels of YUCCA and GH3 genes in myb30 and WT (expression of examined genes in WT was set as 1).Roots of 7-day-old seedlings were used for RNA extraction and RT-qPCR analysis.ACTIN7 gene was used as an internal reference.
Experiments were repeated three times and data are presented as means ± SE (n=3).
Statistical analysis was performed using a two-tailed Student's t-test (*, p < 0.05; **, p < 0.01, compared to WT).F. Relative expression levels of GH3 genes in WT, pi4kγ2, miel1, and myb30 (expression of examined genes in WT was set as 1).Roots of 7-day-old seedlings were used for RNA extraction and RT-qPCR analysis.ACTIN7 gene was used as an internal reference.
Experiments were repeated three times and data are presented as means ± SE (n=3).
Statistical analysis was performed using a two-tailed Student's t-test (*, p < 0.05; **, p < 0.01, compared to WT).G. ChIP-qPCR analysis showed that MYB30 directly binds the promoter regions of GH3.2 and GH3.6 to regulate their expression.Two independent lines expressing MYB30-FLAG (L30 and L36) were used for analysis.Input DNA (from the unimmunoprecipitated DNA) was used as a positive control (100%).
Immunoprecipitated DNA from IgG chromatin was used as a negative control.P1, P2, and P3 are the putative MYB30 binding sequences in promoters of GH3.2 and GH3.6.
C. Relative expression levels of KCS1 and FDH, the target genes for MYB30.WT, pi4kγ2, and miel1 seedlings were used for extracting RNA after inoculation with Pst AvrRpm1 (5х10 7 cfu ml -1 ) for 1 hour.ACTIN7 gene was used as an internal reference.Experiments

Figure 2 .
Figure 2. PI4Kγ2 interacts with E3 ligase MIEL1.A. Yeast two-hybrid assays revealed a direct interaction between PI4Kγ2 and MIEL1.Transformed yeast cells were grown on synthetic drop-out (SD) media (-Leu-Trp) or SD (-Leu-Trp-His-Ade).B.Observation of transiently expressed fusion proteins in Arabidopsis protoplasts showed that PI4Kγ2 colocalizes with MIEL1.Bars=10 μm.C.Bimolecular fluorescence complementation assays confirmed the MIEL1-PI4Kγ2interaction in planta.An Olympus FV1000 confocal microscope was used and collection wavelength was set at 510-550 nm to strictly filter out the autofluorescence (details in "Materials and Methods" section), resulting in that two control panels (upper, middle and right) are blank.Bars=30 μm.

Figure 6 .
Figure 6.PI4Kγ2 is a negative regulator of HR responses in response to bacterial inoculation.A. Symptoms developed by pi4kγ2, miel1, and WT 64 hpi (hour post inoculation) with Pst AvrRpm1 (2х10 6 cfu ml -1 ).Experiments were repeated three times (>5 seedlings each time) and representative images were shown.Scale bar=1 cm.
were repeated three times and data are presented as means ± SE (n = 3).D. A model illustrating the regulation of PI4Kγ2-MIEL1-MYB30 and the effects on plant growth and hypersensitive response (HR).In WT, PI4Kγ2 interacts with MIEL1 to suppress its ubiquitination and degradation, leading to the proteasomal degradation of MYB30 by MIEL1.Reduced MYB30 protein results in increased auxin content/signaling, and in turn promoted growth and suppressed HR to avoid superfluous activation.Under PI4Kγ2 deficiency, enhanced MIEL1 ubiquitination and degradation results in the suppressed ubiquitination and degradation of MYB30.Accumulated MYB30 thereby activates the expression of target genes, resulting in decreased auxin level, suppressed growth, and stimulated HR. References Akhter S, Uddin MN, Jeong IS, Kim DW, Liu XM, Bahk JD (2016) Role of Arabidopsis AtPI4Kγ3, a type II phosphoinositide 4-kinase, in abiotic stress responses and floral transition.Plant Biotechnol J 14: Society of Plant Biologists.All rights reserved.
Society of Plant Biologists.All rights reserved.