Mutations in the TIR1 auxin receptor that increase affinity for auxin/indole-3-acetic acid proteins result in auxin hypersensitivity.

The phytohormone auxin regulates virtually every aspect of plant development. The hormone directly mediates the interaction between the two members of the auxin coreceptor complex, a TRANSPORT INHIBITOR RESPONSE (TIR1)/AUXIN SIGNALING F-BOX protein and an AUXIN/INDOLE-3-ACETIC ACID (Aux/IAA) transcriptional repressor. To learn more about the interaction between these proteins, a mutant screen was performed using the yeast (Saccharomyces cerevisiae) two-hybrid system in Arabidopsis (Arabidopsis thaliana). Two tir1 mutations were identified that increased interaction with Aux/IAAs. The D170E and M473L mutations increase affinity between TIR1 and the degron motif of Aux/IAAs and enhance the activity of the SCF(TIR1) complex. This resulted in faster degradation of Aux/IAAs and increased transcription of auxin-responsive genes in the plant. Plants carrying the pTIR1:tir1 D170E/M473L-Myc transgene exhibit diverse developmental defects during plant growth and display an auxin-hypersensitive phenotype. This work demonstrates that changes in the leucine-rich repeat domain of the TIR1 auxin coreceptor can alter the properties of SCF(TIR1).


INTRODUCTION
The plant hormone indole-3-acetic acid (IAA) is the most important natural auxin with the ability to regulate virtually every aspect of plant development (Woodward and Bartel, 2005;Moller and Weijers, 2009;Sundberg and Ostergaard, 2009;Takahashi et al., 2009;Overvoorde et al., 2010;Vernoux et al., 2010). Auxin signaling is mediated by at least three protein families; the TIR1/AFB F-box proteins, the Aux/IAA transcriptional repressors, and the ARF transcription factors. At low auxin levels, the transcriptional activity of the ARF proteins is inhibited through an interaction with an Aux/IAA protein and the co-repressor TOPLESS (TPL) (Reed, 2001;Tiwari et al., 2001;Weijers et al., 2005;Guilfoyle and Hagen, 2007;Szemenyei et al., 2008). Auxin promotes the recruitment of the Aux/IAA protein to the SCF TIR1/AFB E3 ubiquitin ligase leading to degradation of the Aux/IAA proteins by the ubiquitin-proteasome pathway, thus allowing ARF-dependent transcription (Ruegger et al., 1998;Gray et al., 2001;Mockaitis and Estelle, 2008).
In Arabidopsis, 29 Aux/IAAs have been identified, most of which share a similar protein structure (Reed, 2001). Domain I of the Aux/IAAs binds TPL and is required for transcriptional repression (Tiwari et al., 2004;Szemenyei et al., 2008). The core region of domain II is a 6-amino acid sequence (VGWPPV/I) called the degron, which is required for proteolytic degradation of the Aux/IAA proteins (Worley et al., 2000;Ramos et al., 2001;Dreher et al., 2006). Several dominant or semi-dominant gain-of-function mutants, mutated within the conserved degron sequence, have been isolated (Reed, 2001). Later studies showed that these mutations stabilize the affected Aux/IAAs leading to auxin resistance (Worley et al., 2000;Ramos et al., 2001). Domains III and IV mediate homoand hetero-dimerization, including with the ARF proteins (Mockaitis and Estelle, 2008).
TIR1 is a member of a small family of F-box proteins that contains 5 additional AFB proteins, AFB1-5 (Dharmasiri et al., 2005). All six members function as auxin receptors prepared and incubated with and without auxin. TIR1-Myc was pulled down by anti-c-Myc beads and the amount of AXR3NT-GUS in the complex was determined. The results show that tir1 D170E/M473L-Myc pulls down much more AXR3NT-GUS than the control (Supplemental Fig. 1B). These experiments indicate that the mutations enhance the affinity between TIR1 and Aux/IAA proteins. However, the effect of the mutations can be abolished with anti-auxin compound, auxinole (Hayashi et al., 2012). Auxinole strongly interacts with Phe82 of TIR1, a residue crucial for Aux/IAA recognition, and blocks the formation of the TIR1-IAA-Aux/IAA complex. The result shows that GST-IAA3 cannot pull down additional tir1 D170E/M473L-Myc in the presence of auxinole, similar to the TIR1 control ( Fig. 2B).

D170 and M473 Are Required for the Function of TIR1
The structure of TIR1 shows that D170 and M473 are outside of the auxin binding pocket. D170 is located on the top of the LRRs (Leucine-Rich-Repeats) domain while M473 is in a helix region within the hydrophobic core of this domain (Supplemental Fig.   2A). Interestingly, D170 is conserved among the 6 TIR1/AFB proteins. In addition, the residue corresponding to TIR1 D170 in AFB4 (D250) is mutated to an N in the afb4-2 mutant (Greenham et al., 2011). To determine the effects of the D to N substitution on TIR1, we made this mutant protein and tested it in the pull-down assay. The tir1 D170N protein has reduced interaction with GST-IAA3 (Supplemental Fig. 2B).
To further explore the function of residues D170 and M473, the tir1 mutations D170A and M473A were generated and tested in the two-hybrid system. The results show that DBD-tir1 D170A-Myc displays a reduced interaction with AD-Aux/IAAs in the presence of auxin, while DBD-tir1 M473A-Myc does not interact with AD-Aux/IAAs compared to the control (Fig. 2C). Similar results were obtained by in vitro pull-down assay. The D170A substitution reduced recovery of TIR1 while the M473A-Myc protein did not respond to auxin (Fig. 2D). These results indicate that D170 and M473 are required for TIR1 function.

The TIR1 Mutants Enhance the Interaction with the Degron Motif of Aux/IAAs
Structural studies have established the spatial relationship between TIR1 and the degron motif of IAA7. However, the structure of other domains of IAA7 and their potential interaction with TIR1 is not known. To further study the effect of D170M and M473L, the interaction between the TIR1 mutants and different fragments of IAA7 was determined by yeast two-hybrid test. The results show that the mutations increase the interaction between TIR1 and different IAA7 fragments including DI/II, DI/II/III and DII/III/IV, suggesting that the mutations may enhance the interaction with DII of IAA7 (Fig. 3A). To address this possibility, a yeast 2-hybrid experiment was performed to test the interaction between the degron motif from IAA7 and the tir1 D170E M473L. The results show that the mutant protein interacts better with domain II of IAA7 upon auxin treatment than does the control ( Fig. 3B and 3C). Similar results were observed in pull-down assays. The tir1 D170E/M473L protein does not interact with GST-DI or DIII/IV of IAA7 (Supplemental Fig.   1C). However, GST-degron motif pulls down much more tir1 D170E/M473L-Myc than the TIR1-Myc control in the absence and presence of auxin (Fig. 3D). This indicates that the TIR1 mutations enhance the interaction with the degron motif of Aux/IAAs and that this effect is at least partially independent of auxin.
To determine whether the effect of the mutations is specific to IAA, different auxins were tested, including 4-Cl-IAA, NAA, 2,4-D and picloram. The results show that each of these auxins except picloram dramatically increased the interaction. The behavior of picloram is consistent with earlier studies showing that TIR1 does not recognize picloram (Calderon Villalobos et al., 2012) (Supplemental Fig. 1D).

The TIR1 Mutations D170E and M473L Enhance the Degradation of Aux/IAAs
Although neither D170 nor M473 are in the F-box domain, it is possible that the mutations affect the interaction with the ASK1 adaptor protein. We tested this possibility in 1 0 yeast. The results show that the TIR1 mutants exhibited a similar interaction with ASK1 as the control, suggesting that the mutations do not affect the formation of the TIR1-ASK complex (Supplemental Fig. 3).
To explore the effects of the mutations on degradation of Aux/IAAs, we used the pTIR1:tir1 D170E/M473L-Myc pHS:AXR3NT-GUS transgenic line. GUS activity was measured after heat shock treatment. Compared to the control, the tir1 D170E/M473L-Myc transgenic seedlings display lower GUS activities after heat shock treatment in the absence or presence of auxin, suggesting that the tir1 D170E/M473L-Myc transgenic plants degrade AXR3NT faster than the control ( Fig. 4A and 4B). A protein blot showed that the transgenic lines expressed TIR1 at a similar level (Fig. 4C).
As a complement to this experiment we also assessed the effects of the mutation on Aux/IAA degradation in Saccharomyces cerevisiae (Havens et al., 2012). Since eukaryotes share conserved components in the SCF-dependent degradation pathway, YFP-Aux/IAA fusion proteins are rapidly degraded by SCF TIR1 complex upon auxin treatment in yeast (Nishimura et al., 2009;Havens et al., 2012) Fig. 4).
We also examined degradation of two truncated IAA28 proteins containing DII; t2, centered on the DII domain, and t2v, shifted N-terminal to include the upstream KR 1 1 sequence (Fig 5B and Table 1). Both proteins were rapidly degraded by D170E M473L, consistent with our two-hybrid and pulldown assays, indicating that the mutations affect the interaction with the DII domain. Finally, we examined degradation of IAA20, which lacks the DII domain. This protein is stable confirming that the DII region is required for degradation by either wild-type or mutant proteins (Fig. 5B).

Exhibit an Auxin-hypersensitive Phonotype
To explore the effect of the tir1 mutations on plant development, the phenotype of two To characterize the auxin response in these lines, the effect of auxin on root growth was determined. 6-day old seedlings were transferred onto fresh ATS medium with different concentrations of the synthetic auxin 2,4-D. After another two days of growth, the length of newly grown primary root was measured and expressed relative to growth on control plates. The result shows that the primary root elongation is strongly inhibited in pTIR1:tir1 D170E/M473L-Myc transgenic seedlings at low concentrations of 2,4-D compared with the control (Fig. 7A). This suggests that tir1 D170E/M473L causes an auxin-hypersensitive phenotype in the plant.
Next we measured the expression of auxin responsive genes in the tir1 D170E/M473L-Myc transgenic lines using real time Q-PCR. Our results show that expression of the tir1 D170E/M473L proteins increase transcription of auxin responsive genes in the plant. For example, the tir1 D170E/M473L-Myc transgenic seedlings display a significantly increased IAA3 RNA level compared with the wild type TIR1 control with or 1 2 without 2,4-D treatment (Fig. 7B). Similar effects can be observed for other auxin responsive genes as well (Fig. 7B).

The Function of the Mutations May be Unique for the TIR1 Protein
TIR1 belongs to a small family of F-box proteins that contains 5 additional AFB proteins, AFB1-5 (Dharmasiri et al., 2005). As previously noted, D170 is conserved in all 6 proteins, and the afb4-2 mutation affects this residue (Greenham et al., 2011). However we find that the corresponding D to E substitution in AFB1 and AFB2 does not increase binding to the Aux/IAAs (See Supplementary Fig. 5B and 5C). In the case of AFB2, the mutation decreased the interaction. Either M or L can be found in other family members at positions homologous to TIR1 M473. The M468L mutation in AFB2 does not affect interaction with the Aux/IAAs. Thus, the effect of TIR1 D170E and M473L are dependent on the TIR1 context.

DISCUSSION
The ubiquitin-proteasome pathway is one of the most important proteolytic pathways in eukaryotes. In this pathway, the small protein ubiquitin (Ub) is attached to protein substrates and the Ub-protein conjugates are recognized and degraded by the 26S proteasome (Fang and Weissman, 2004). The SCF complexes are a major class of ubiquitin ligase enzyme (Gagne et al., 2002;Petroski and Deshaies, 2005). In this complex, the F-box protein plays a critical role in the determination of substrate specificity (Petroski and Deshaies, 2005). The plant hormone auxin directly induces rapid degradation of the Aux/IAA family of transcriptional repressors by SCF TIR1/AFB E3 ubiquitin ligase (Gray et al., 2001;Dharmasiri et al., 2005;Dharmasiri et al., 2005;Kepinski and Leyser, 2005;Tan et al., 2007). In Arabidopsis, TIR1 is a member of a small group of F-box proteins that also includes AFB1 through AFB5 and the jasmonic acid receptor, COI1 (Dharmasiri et al., 2005). The TIR1 protein plays a critical role in regulating most aspects 1 3 of auxin response.
In this study, two TIR1 mutations, D170E and M473L, were isolated that increase the interaction between TIR1 and the Aux/IAA proteins ( Fig. 1A and 1C). Further studies show that the TIR1 mutations increase the affinity for the degron motif of the Aux/IAAs ( Fig. 2A and Supplemental Fig. 1B). Both D170 and M473 are located outside of the TIR1 auxin-binding pocket and do not directly contact auxin or the degron motif of Aux/IAAs (Tan et al., 2007). However, when they are mutated to A, they either reduce the interaction with IAA7 for D170A, or abolish the interaction in the case of M473A, indicating that these two positions are essential for TIR1 function ( Previous studies demonstrated that the TIR1-IAA-degron module is a powerful tool to study protein function. The auxin-inducible degron (AID) system can be used to control the degradation of target proteins upon auxin treatment in a reversible manner in nonplant cells such as yeast, chicken, mouse, hamster, monkey and human cells (Nishimura et al., 2009;Kanke et al., 2011;Holland et al., 2012). The system is particularly useful for investigating the function of proteins in specific developmental processes (Holland et al., 2012). In our yeast analysis, the tir1 D170E/M473L protein increases the rate of degradation of YFP-Aux/IAA7/28 approximately 15 to 25 fold compared to wild-type TIR1 ( Fig. 6). Our results suggest that it may be possible to modify the TIR1/AFB proteins from Arabidopsis or other plant species to increase the flexibility of the AID system.

Generation of mutant library and mutant screen
Full length TIR1 cDNA was mutagenized by error-prone PCR with a mutation ratio of approximately 2 mutations per tir1 molecule. The PCR products were ligated into the pGILDA vector and transformed into E. coli competent cells to generate a library of approximately 1.3×10 4 colonies. The library was transformed into the pB42AD:IAA12 yeast strain (EGY48) and yeast colonies were screened on SD-U-H-W-L medium supplemented with 10μM IAA. The number of total yeast colonies screened was about 2.5×10 5 . Yeast colonies that grew fastest were isolated and tested on X-gal plates. The pGILDA:tir1-Myc plasmids from positive colonies were extracted and sequenced.

Plant materials and conditions
All Arabidopsis thaliana mutants used in this study were generated in the Columbia-0 (Col-0) ecotype. Seeds were surface sterilized for 20min in 30% commercial bleach, www.plantphysiol.org on August 18, 2017 -Published by Downloaded from Copyright © 2013 American Society of Plant Biologists. All rights reserved. plated on ATS medium (Arabidopsis thaliana solution) supplemented with 0.8% agar, and stratified for 2 -4 days at 4°C. ATS medium consists of 1% sucrose, 5mM KNO 3 , 2.5mM KPO 4 , 2mM MgSO 4 , 2mM Ca(NO 3 ) 2 , 50μM Fe-EDTA, and 1mL/L of micronutrients. All seedling experiments were performed under long day conditions (16h light and 8h dark) a growth chamber (80μmol/m 2 /s, 22°C), unless otherwise stated. Plants in soil were grown in long day conditions at 22°C.

Plasmid constructs and generation of transgenic lines
The pTIR1:TIR1/tir1-Myc constructs were made by introducing a 2kb 5' upstream region of the TIR1 gene adjacent to the TIR1/tir1 cDNA into pGWB16 vector. In all cases, plasmids were introduced into Col-0 plants as described in (Clough and Bent, 1998)

RNA extraction and real time RT-PCR
Total RNA was extracted from 7-day-old seedlings by the RNeasy plant mini kit (Qiagen) and RNA yield was quantified using a Thermo Scientific NanoDrop 2000. 1μg of RNA was used for cDNA synthesis using the Superscript III First-Strand Synthesis kit (Invitrogen). Quantitative RT-PCR was performed as previously described (Greenham et al., 2011). Data was normalized to the reference PP2AA3 according to the △ Ct method.

Protein extraction and pull-down assays
To perform pull-down assays from plant extracts, tissue was harvested, ground to a powder in liquid nitrogen, and vortexed vigorously in extraction buffer (50mM Tris pH7.5, 150mM NaCl, 10% glycerol, 0.1% NP-40, complete protease inhibitor (Roche), 20μM MG132). Cellular debris was removed by centrifugation and total protein concentration was determined using the Bradford assay. Total protein extract (1mg) was incubated with or without 50μM IAA at 4°C for 4hr. TIR1-Myc was recovered with 20μl of anti-c-Myc agarose beads (clontech) and washed by extraction buffer for 5 times. The protein 1 6 sample was eluted by SDS-PAGE sample buffer, heated for 8min at 90°C, cooled on ice for 2min and fractionated by SDS-PAGE. The AXR3NT-GUS protein was detected by immunoblotting with anti-GUS (Invitrogen) and visualized using the ECL Plus Western Blotting Detection System (Amersham).
The in vitro pull-down assay from TNT® coupled wheat germ extracts was described previously (Parry et al., 2009). Briefly, GST-Aux/IAA was expressed in E.coli and purified with glutathione agarose beads (Sigma). TIR1/tir1-Myc proteins were synthesized in the TNT® coupled wheat germ extract system (Promega). 20μl of extract was incubated with GUS staining was performed on 7-day old seedling. The seedlings were collected in GUS staining solution (100mM Na 2 PO 4 pH 7.0, 10mM EDTA, 0.1% Triton X-100, 1.0mM K 3 Fe(CN) 6 and 2mM X-Gluc), vacuum-infiltrated for 20 min and stained overnight at 37°C.
The seedlings were cleared in 70% (v/v) ethanol and imaged with a Nikon SMZ1500 dissecting microscope. MUG assays were performed to quantify β -glucuronidase activity (Ge et al., 2010). To measure β -galactosidase activity, proteins were extracted from yeast cells using Y-PER Reagent (Thermo) and the yield was determed by Bradford assay. The assay was performed using 100μl protein extract, 200μl 4mg/ml ONPG with 700μl

Root inhibition assay
For root growth assays, 6-day old seedlings were transferred onto fresh ATS medium with different concentrations of 2,4-D for 2 days and the length of new primary root was measured using ImageJ software.

Aux/IAA Degradation Assays in Yeast
Yeast degradation assays were carried out as in (Havens et al., 2012). Briefly, yeast strains co-expressing stably-integrated TIR1 variants and YFP-tagged IAA proteins were prepared by transferring a freshly grown colony from YPD plates into Synthetic Complete (SC) media. Flow cytometry was used to estimate the cell density (in eventsμL −1 ) and dilute cells to such that cultures were in log phase 16 h later and for the duration of the experiment. All cultures were grown at 30°C with shaking. Pre-auxin measurements were taken to ascertain baseline expression, followed by addition of auxin (10μM indole-3-acetic acid) or mock treatment (95% [v/v] ethanol). Measurements were acquired over the course of 120-150 min following auxin treatment, with intervals ranging from 3 min early in the YFP-IAA degradation phase to 20 min later in the degradation phase. Controls were measured every hour for the duration of the experiment.

Quantitative Analysis of IAA Degradation Rate
Quantitative analysis of IAA degradation profiles obtained in yeast was conducted as in (Havens et al., 2012). Briefly, the auxin-induced degradation dynamics of YFP-Aux/IAA fusion proteins was characterized by a second-order differential equation model, x' = k1 u -k2 x y' = k3 -k4 y -k5 x y. group, two parameters, k3 and k5, approximate the expression and degradation rates of an Aux/IAA protein respectively and were allowed to vary in the estimation process while holding the rest of the parameters constant. This global fitting approach ensured that the degradation variability among the unique TIR1 and YFP-Aux/IAA pairs are represented in the estimated k5 values. The model fit residual was computed using 2-norm of the difference between the measured and model predicted YFP intensity at the measurement times 0 <= t <= 150 min.

Accession Numbers
Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: TIR1 (At3G62980), r  e  n  c  e  s   C  a  l  d  e  r  o  n  V  i  l  l  a  l  o  b  o  s  L  I  ,  L  e  e  S  ,  D  e  O  l  i  v  e  i  r  a  C  ,  I  v  e  t  a  c  A  ,  B  r  a  n  d  t  W  ,  A  r  m  i  t  a  g  e  L  ,  S  h  e  a  r  d  L  B  ,  T  a  n  X  ,  P  a  r  r  y  G  ,  M  a Figure 1. Identification of TIR1 mutants with increased binding to the Aux/IAA proteins.
(A) The TIR1 mutations D170E and M473L increase the interaction between TIR1 and different Aux/IAAs upon auxin treatment. (B) The mutant proteins do not interact with degron-mutated IAA7 proteins. The degron motif of IAA7 is VGWPPV, while it is mutated to VGWSPV in iaa7-1; VGWPSV in iaa7-2 and VGWSSV in iaa7-3. (C) Measurement of β-galactosidase activity in yeast strains with and without auxin treatment.
The asterisks indicate statistically significant differences between the mutants and the wild type control (t-test, P < 0.01). Error bars are SEM.  (A) Y2H interaction between TIR1 single and double mutant proteins with IAA7 fragments. (B) Interaction between TIR1 D170E/M473L and the IAA7 degron. (C) β-galactosidase activity in yeast cells expressing the mutants and the IAA7 degron. Error bars are SEM. The asterisks denote statistically significant differences between the mutants and the wild-type control (t-test P < 0.01). (D) In vitro pull-down assay shows that GST-degron motif pulls down more tir1 D170E/M473L-Myc protein than wild-type TIR1-Myc. The degron motif is from position 83 to 92 of IAA7. Numbers indicate the fold changes relative to the control sample.