Regulation of Arabidopsis SHY2/IAA3 protein turnover


For correspondence (fax +1 919 962 1625; e-mail


Auxin/indole acetic acid (Aux/IAA) proteins regulate transcriptional responses to the plant hormone auxin. Gain-of-function mutations in the Arabidopsis SHORT HYPOCOTYL 2 (SHY2/IAA3) gene encoding an Aux/IAA protein increase steady-state levels of SHY2/IAA3 protein and decrease auxin responses, indicating that SHY2/IAA3 negatively regulates auxin signaling. These shy2 mutations also cause ectopic light responses, suggesting that SHY2/IAA3 may promote light signaling. Auxin regulates turnover of the related Auxin-resistant (AXR)2/IAA7 and AXR3/IAA17 proteins by increasing their interaction with the Skp1-Cdc53/cullin-F-box (SCFTIR1) E3 ubiquitin ligase complex. To investigate whether SHY2/IAA3 is regulated similarly, we have used a turnover assay to reveal that axr1 and transport inhibitor resistant (tir)1 mutations affecting SCFTIR1 decrease SHY2/IAA3 turnover. In pull-down assays, SHY2/IAA3 protein interacted with TIR1, the F-box component of SCFTIR1 and with the photoreceptor phytochrome B. Auxin stimulated SHY2/IAA3 interaction with TIR1, whereas the shy2-2 gain-of-function mutation decreased this interaction. Light did not affect the interaction, suggesting that light regulates some other aspect of Aux/IAA gene or protein function. The chemical juglone (5-hydroxy-1,4-naphthoquinone) inhibited the interaction, suggesting that peptidyl-prolyl isomerization may mediate auxin-induced SHY2/IAA3 protein turnover.


The plant hormone auxin induces or represses numerous genes, and this regulation is mediated by auxin response factors (ARFs) that bind to auxin response promoter elements upstream of these genes (Hagen and Guilfoyle, 2002). In addition to an N-terminal DNA-binding domain, ARF proteins have a middle region that can either activate or repress gene expression in transient assays, and C-terminal sequences called domains III and IV that can mediate dimerization (Kim et al., 1997; Ouellet et al., 2001; Ulmasov et al., 1999). Domains III and IV are required for auxin responsiveness of ARF proteins (Tiwari et al., 2003). Arabidopsis has at least 22 ARF genes that probably mediate diverse aspects of gene expression responses to auxin (Hagen and Guilfoyle, 2002).

A family of 29 indole acetic acid (IAA) genes in Arabidopsis encode Aux/IAA proteins, named for the auxin inducibility of the first such genes discovered. Aux/IAA proteins have domains III and IV similar to those of ARF proteins and can dimerize with ARF proteins in yeast two-hybrid assays through these domains (Kim et al., 1997; Ulmasov et al., 1997). Overexpression of IAA genes in transient transfection assays and characterization of gain-of-function mutations in several IAA genes indicate that Aux/IAA proteins can inhibit auxin gene expression responses (Abel et al., 1995; Tian et al., 2002; Tiwari et al., 2001; Ulmasov et al., 1997), probably by dimerizing with ARF proteins and modulating their activity. Many IAA genes are themselves regulated by auxin (Abel et al., 1995), suggesting that feedback loops may modulate temporal or spatial patterns of auxin-regulated gene expression.

Gain-of-function mutations in any of several Arabidopsis IAA genes changed amino acids in conserved domain II and caused auxin-related phenotypes (Reed, 2001). Among these, the short hypocotyl 2 (shy2-1), shy2-2, and shy2-3 mutations in SHY2/IAA3 caused short hypocotyls, slightly auxin resistant hypocotyl and root growth, reduced lateral root formation or outgrowth, slowed gravitropic response, and decreased auxin-regulated gene expression responses (Kim et al., 1996, 1998; Oono et al., 2002; Reed et al., 1998; Soh et al., 1999; Tian and Reed, 1999; Tian et al., 2002). shy2-2 plants have increased level of SHY2/IAA3 protein (Colón-Carmona et al., 2000), suggesting that these mutations stabilize SHY2/IAA3, thereby allowing it to inhibit auxin responses. Domain II mutations in several other IAA genes also stabilized the corresponding proteins or fusion proteins to GUS or luciferase (Gray et al., 2001; Ouellet et al., 2001; Ramos et al., 2001; Tiwari et al., 2001; Worley et al., 2000).

Light regulates expression of SHY2/IAA3 and other Aux/IAA genes (Tepperman et al., 2001; Tian et al., 2002), and shy2-2 mutants make leaves and overexpress several light-regulated genes in the dark (Kim et al., 1996, 1998; Reed et al., 1998; Soh et al., 1999; Tian and Reed, 1999; Tian et al., 2002). Similar mutants in AXR2/IAA7 and AXR3/IAA17 also made leaves in darkness (Nagpal et al., 2000). Moreover, SHY2/IAA3 and other Aux/IAA proteins can interact with the photoreceptor phytochrome A in pull-down assays, and oat phytochrome A can phosphorylate SHY2/IAA3 protein in vitro (Colón-Carmona et al., 2000). These results have suggested that light might activate SHY2/IAA3 and other Aux/IAA proteins by stabilizing them, which might in turn activate photomorphogenesis.

Other data suggest that auxin regulates turnover of Aux/IAA proteins. Auxin increased turnover of AXR2/IAA7::GUS, AXR3/IAA17::GUS, Pisum sativum (PS)-IAA6::luciferase (LUC), and IAA1::LUC fusion proteins in intact plants, and auxin promoted interactions between AXR2/IAA7 and AXR3/IAA17 fusion proteins and the SCFTIR1 E3 ubiquitin ligase complex (Gray et al., 2001; Zenser et al., 2001). SCFTIR1 has four protein subunits encoded by the cullin (CUL)1, Arabidopsis (ASK)1, ring box (RBX)1, and TIR1 genes. TIR1 is an F-box protein that determines SCFTIR1 substrate specificity. Moreover, mutations in CUL1, TIR1, or ASK1, encoding components of SCFTIR1, or in AXR1, required for modification of the cullin component of SCFTIR1 with the ubiquitin-related protein Rub, all reduced auxin responses (Hellmann et al., 2003; Ward and Estelle, 2001).

To determine whether SCFTIR1 also regulates SHY2/IAA3 protein turnover, we used recombinant SHY2/IAA3 protein to assay turnover in plant extracts of various genotypes. We also used a pull-down assay to detect interaction between SHY2/IAA3 and TIR1, the F-box component of SCFTIR1 and between SHY2/IAA3 and phytochrome B. We have used these assays to explore regulation of SHY2/IAA3 protein turnover by various known genes, as well as by auxin and light.


Expression of SHY2/IAA3 protein and antibody production

We expressed SHY2/IAA3::intein-chitin binding domain (CBD) and maltose binding protein (MBP)::SHY2/IAA3 fusion proteins in Escherichia coli (Figure 1a), and used purified SHY2/IAA3 protein cleaved from SHY2/IAA3::intein-CBD to raise anti-SHY2/IAA3 polyclonal antibody. The immune, but not the pre-immune serum, recognized the purified SHY2/IAA3 protein and both fusion proteins, but did not recognize any protein in crude extracts from E. coli that was not expressing SHY2/IAA3 (Figure 1b,c; data not shown). Aux/IAA proteins share conserved domains, and this antiserum also recognized AXR2/IAA7::GST fusion protein (Gray et al., 2001; Figure 1d).

Figure 1.

Expression of SHY2/IAA3 and shy2-2 fusion proteins in E. coli and anti-SHY2/IAA3 antibody production.

(a) Coomassie-stained gels of extracts from E. coli cells expressing SHY2/IAA3::intein-CBD, MBP::SHY2/IAA3, intein-CBD, or MBP proteins, in the absence (−) or presence (+) of IPTG. Arrowheads show the positions of the full-length fusion proteins, and the asterisks indicate the positions of intein-CBD or MBP proteins.

(b) Immunoblots of replicas of gels shown in (a), probed with anti-SHY2/IAA3 antibody.

(c) Immunoblot of purified SHY2/IAA3 and shy2-2 proteins (arrowhead) probed with anti-SHY2/IAA3 antibody. Each lane contains 20 ng protein.

(d) Recognition of AXR2/IAA7::GST by anti-SHY2/IAA3 antibody. Extracts from non-induced (−IPTG) or induced (+IPTG) E. coli cells expressing AXR2/IAA7::GST fusion protein or GST alone were separated by SDS–PAGE, blotted, and probed with anti-SHY2/IAA3 antibody. The arrowhead indicates the position of AXR2/IAA7::GST fusion protein.

The affinity-purified antiserum could detect as little as 10 ng SHY2/IAA3 protein expressed in E. coli (Figure 1c, data not shown). When 150 µg of total protein extract from wild-type Arabidopsis seedlings was immunoblotted, the affinity-purified antiserum detected several proteins including one with the same size as the E. coli expressed SHY2/IAA3 protein. However, this band was also present in extracts from the SHY2/IAA3 null mutant shy2-24 (data not shown). Immunoprecipitations from 35S-labeled seedlings also revealed a band of the predicted size that was still present in shy2-24 extracts. Therefore, this antiserum could recognize a protein of the same size as SHY2/IAA3, possibly another closely related Aux/IAA protein. We also failed to detect any increased level of SHY2/IAA3 protein in extracts of 35S::SHY2/IAA3 or shy2-2 plants, suggesting that the protein is present at a very low level. This is consistent with data from previous reports (Abel et al., 1994; Colón-Carmona et al., 2000).

AXR1 and TIR1 promote SHY2/IAA3 turnover

Low SHY2/IAA3 protein level made in vivo studies of SHY2/IAA3 turnover impractical. To characterize SHY2/IAA3 protein turnover, we therefore developed an in vitro assay. Purified SHY2/IAA3 protein from E. coli was incubated with plant crude extracts for different periods of time, and then examined by immunoblots. As shown in Figure 2, when 0.3 µg of SHY2/IAA3 protein was incubated with 200 µg total protein extracts from wild-type Arabidopsis seedlings at 4°C, the SHY2/IAA3 protein disappeared over the course of several hours. The 26S proteasome inhibitor MG132 (Rock et al., 1994) inhibited this disappearance (Figure 2a), suggesting that SHY2/IAA3 protein is degraded by the 26S proteasome in this assay.

Figure 2.

Turnover of SHY2/IAA3 and shy2-2 proteins in plant extracts.

Purified SHY2/IAA3 protein (0.3 µg) cleaved from SHY2/IAA3::intein-CBD was incubated with 200 µg crude plant extracts of the indicated genotypes. Aliquots were removed after the indicated incubation times, separated by SDS–PAGE, blotted, and detected with anti-SHY2/IAA3 antibody.

(a) Time course of degradation of SHY2/IAA3 protein by extracts from 20-day-old wild-type Arabidopsis seedlings, in the absence (1% DMSO) or presence of 100 µm MG132. Ten repetitions of this experiment (with some variations in relative amounts of reaction components) gave similar results.

(b) Time course of turnover of SHY2/IAA3 protein added to extracts from wild-type, axr1-12, or (c) tir1-1 seedlings. Duplicate wild-type samples were used in the two blots. This experiment was repeated eight times for axr1-12 and five times for tir1-1 with similar results.

To determine whether the SCFTIR1 ubiquitin ligase might regulate SHY2/IAA3 protein turnover, we assayed SHY2/IAA3 turnover in crude extracts from the strong axr1-12 and tir1-1 mutants. As shown in Figure 2(b), SHY2/IAA3 was more stable in axr1-12 extracts than in wild-type extracts. SHY2/IAA3 protein was also more stable in tir1-1 mutant extracts than in the wild-type extracts, although it was still degraded to some degree (Figure 2c).

We used this assay to search for other genes that might regulate SHY2/IAA3 protein turnover. We tested the turnover of SHY2/IAA3 in extracts from mutant or transgenic plants affected in several genes possibly involved in auxin homeostasis, auxin response, light response, or protein turnover. These included phyB-1 (Reed et al., 1993), suppressor of axr1 (sar)1-1 (Cernac et al., 1997), sar1-1 axr1-3 (Cernac et al., 1997), axr6-2 (Hobbie et al., 2000), ask1-1 (Yang et al., 1999), hookless1-1 (Lehman et al., 1996), pinoid (pin)-3 (Bennett et al., 1995), 35S::GH3 (Tom Guilfoyle, unpublished), 35S::iaaL (Jensen et al., 1998; Romano et al., 1991), hobbit-2311, hobbit-5423, hobbit-5859 (Willemsen et al., 1998), COP9 signalosome (CSN)-5 antisense (Schwechheimer et al., 2001), fusca (fus)5-T379, fus11-U203, fus6-1, constitutively photomorphogenic (cop)1-4, cop1-8, de-etiolated (det)1-1, and 35S::COP1 (McNellis et al., 1994; Wei and Deng, 1996). SHY2/IAA3 protein disappeared in these extracts at the same rate as in wild-type extracts (data not shown), suggesting that those genes are probably not rate-limiting for the degradation of SHY2/IAA3, at least in our assay conditions.

Auxin promotes SHY2/IAA3 interaction with TIR1::myc

We explored direct interactions of SHY2 with the SCFTIR1 ubiquitin ligase using a pull-down assay. Chitin beads loaded with SHY2/IAA3::intein-CBD fusion protein were incubated with total protein extracts from tir1-1 seedlings expressing c-myc epitope-tagged TIR1 (Gray et al., 1999) and pulled down, and a gel of the precipitated proteins was immunoblotted with anti-myc antibody. As shown in Figure 3(a), TIR1::myc protein co-precipitated with SHY2/IAA3::intein-CBD fusion protein. It did not precipitate with the CBD alone (data not shown).

Figure 3.

Pull-down of TIR1::myc by SHY2/IAA3::intein-CBD.

Chitin beads loaded with SHY2/IAA3::intein-CBD or shy2-2::intein-CBD protein were added to extracts of tir1-1[TIR1::myc] plants, and the precipitated proteins were separated by SDS–PAGE, blotted, and detected using anti-c-myc antibody. The arrows indicate the position of TIR1::myc protein, and the asterisk indicates the position of a non-specific band which serves as an input control for SHY2/IAA3 or shy2-2 fusion proteins.

(a) SHY2/IAA3::intein-CBD and shy2-2::intein-CBD fusion proteins were used in pull-down assays with the extracts from either tir1-1[TIR1::myc] (TIR1wt) or tir1-1[TIR1P10A::myc] (TIR1mt) seedlings.

(b) Pull-down assays with SHY2/IAA3::intein-CBD fusion protein were performed in the absence (none) or presence of IAA, cycloheximide (cyclo), or both IAA and cycloheximide as described in Experimental procedures.

(c) SHY2/IAA3::intein-CBD and shy2-2::intein-CBD proteins were incubated with tir1-1[TIR1::myc] extracts in the absence (none) or presence of 2 or 20 µm IAA for 3 or 6 h before proteins were pulled down.

(d) Dark-grown tir1-1[TIR1::myc] seedlings were given a pulse of red light for 10 min and returned to darkness for 2 h (D/R/D), or grown continuously in the dark (D), or treated with 20 µm IAA for 2 h (D/+IAA) before extracting proteins for the pull-down assay.

We also probed blots from these pull-down experiments with anti-CUL1 and anti-ASK1 antibodies, but failed to observe a signal in the precipitates (data not shown). This result suggests that TIR1 can bind SHY2/IAA3 without being part of an SCF E3 complex, or that some aspect of our assay conditions prevented our detecting the interaction. Consistent with the former possibility, the P10A mutation in the F-box of TIR1::myc, which decreases TIR1 affinity for ASK1 and ASK2 (Gray et al., 1999), did not affect the TIR1::myc–SHY2/IAA3 interaction (Figure 3a).

SHY2/IAA3 protein interacted much more efficiently with TIR1::myc protein from extracts of auxin-treated seedlings than it did with TIR1::myc from extracts of untreated seedlings (Figure 3b). Treating the seedlings with the protein synthesis inhibitor cycloheximide did not affect this interaction in either the absence or the presence of auxin (Figure 3b). Auxin did not affect the amount of TIR1::myc present in the plants (Figure 3b). Moreover, auxin treatment of extracts also promoted the interaction between SHY2/IAA3 and TIR1::myc (Figure 3c). Together, these results indicate that auxin promotes the TIR1::myc–SHY2/IAA3 interaction by a post-translational mechanism. We also tested whether auxin affected disappearance of SHY2/IAA3 protein in our turnover assay, but failed to see a consistent effect (data not shown).

The shy2-2 mutation decreases SHY2/IAA3 interaction with TIR1::myc

In the pull-down assay, much less TIR1::myc protein co-precipitated with shy2-2::intein-CBD fusion protein than with wild-type SHY2/IAA3::intein-CBD fusion protein (Figure 3a). These results suggest that the shy2-2 mutation stabilizes the protein by decreasing its interaction with TIR1 and perhaps other F-box proteins. However, the shy2-2 mutation did not eliminate auxin regulation of this interaction. Inclusion of auxin in the pull-down assay increased the amount of TIR1::myc that shy2-2::intein-CBD fusion protein pulled down, although SHY2/IAA3::intein-CBD fusion protein always pulled down more TIR1::myc than did shy2-2::intein-CBD under the same conditions (Figure 3c). In the turnover assay, we did not see a difference in rate of disappearance of shy2-2 protein compared to wild-type SHY2/IAA3 protein. This observation suggests that the pull-down assay is more sensitive to quantitative differences in interaction with TIR1 than is the turnover assay.

Juglone inhibits the interaction between TIR1::myc and SHY2/IAA3

The TIR1::myc pull-down assay provided a means to explore potential mechanisms by which auxin might regulate recognition of SHY2/IAA3 by TIR1. We tested whether various pharmacological agents might affect the interaction. We found that juglone (5-hydroxy-1,4-naphthoquinone), an inhibitor of the parvulin class of peptidyl-prolyl isomerases (Hennig et al., 1998), inhibited the TIR1::myc interaction in both the absence and the presence of auxin (Figure 4). Several other agents had no effect (Figure 4 and data not shown), including curcumin, an inhibitor of a COP9 signalosome-associated kinase (Sun et al., 2002), 3,4-dehydro-proline, an inhibitor of proline hydroxylation (Ivan et al., 2001), the kinase inhibitor staurosporine, and the phosphatase inhibitors okadaic acid, cantharadin, and NaF. Treatment with calf intestinal phosphatase also had no effect on the interaction.

Figure 4.

Juglone inhibits interactions between TIR1::myc and SHY2::intein-CBD.

Chitin beads loaded with SHY2/IAA3::intein-CBD were incubated with extracts from tir1-1[TIR1::myc] seedlings in the presence of various chemicals, and the precipitated proteins were separated by SDS–PAGE, blotted, and detected using anti-c-myc antibody. The treatments were as follows: lane 1, no treatment; lane 2, 10 µm curcumin; lane 3, 50 µm juglone; lane 4, 100 µm 3,4-dehydro-l-proline; lane 5, 20 µm IAA; lane 6, 20 µm IAA + 10 µm curcumin; lane 7, 20 µm IAA + 50 µm curcumin; lane 8, 20 µm IAA + 50 µm juglone; lane 9, 20 µm IAA + 100 µm 3,4-dehydro-l-proline. The arrow indicates the position of TIR1::myc protein.

Phytochrome B can interact with SHY2/IAA3

The photoreceptor phytochrome A from oat can interact with various Aux/IAA proteins including SHY2/IAA3 (Colón-Carmona et al., 2000). We tested whether phytochrome B could also interact with SHY2/IAA3 in pull-down assays with SHY2/IAA3::intein-CBD-loaded chitin beads. SHY2/IAA3::intein-CBD effectively pulled down a phyB::myc fusion protein (Reed et al., 2000) as well as native phyB protein from plant extracts (Figure 5a,b). The shy2-2 mutant protein interected with phyB as efficiently as did the wild-type SHY2/IAA3 protein, as indicated in a dilution series using different amounts of SHY2/IAA3::intein-CBD or shy2-2::intein-CBD fusion proteins (Figure 5b). A fusion to AXR2/IAA7 protein (Gray et al., 2001) also interacted efficiently with phyB (L. Krall and J. W. R., unpublished result), suggesting that multiple Aux/IAA proteins can interact with phyB as well as with phyA. The SHY2/IAA3–phyB interaction was not affected by light, as in the case of the interaction between oat phyA and Aux/IAA proteins (Colón-Carmona et al., 2000). Light might, nevertheless, regulate the interaction in vivo, as it is known that light promotes translocation of phytochromes to the nucleus where Aux/IAA proteins reside.

Figure 5.

Pull-down of PHYB protein by SHY2::intein-CBD.

Chitin beads loaded with SHY2/IAA3::intein-CBD or shy2-2::intein-CBD protein were added to extracts from indicated genotypes, and the precipitated proteins were separated by SDS–PAGE, blotted, and probed using indicated antibodies.

(a) PHYB::myc protein detected with anti-myc antibody. S, supernatant; W, last wash from precipitated beads; P, pellet after washing.

(b) phyB pulled down from extracts of 35S::PHYB plants with serial dilutions of chitin beads loaded with either SHY2/IAA3::intein-CBD or shy2-2::intein-CBD, and detected with anti-phyB monoclonal antibody (Shinomura et al., 1996) or anti-SHY2/IAA3 antibody.

(c,d) phyB pulled down from extracts of plants overexpressing wild type or mutant versions of phyB with SHY2/IAA3::intein-CBD or shy2-2::intein-CBD, and detected with anti-phyB antibody. phyB-9 is a null mutant control. Overexpressed phyB mutant genes have been described previously by Krall and Reed (2000) and Wagner and Quail (1995).

We tested whether several available mutant versions of phyB could interact with SHY2/IAA3, including several point mutations in the Per-Arnt-Sim (PAS) repeat domain (Wagner and Quail, 1995), and a truncation of part of the C-terminal histidine kinase-related domain (phyB-28) (Krall and Reed, 2000). None of these mutations eliminated the interaction (Figure 5c,d), suggesting that the interaction does not require these C-terminal portions of phyB.

The interactions of SHY2/IAA3 and other Aux/IAA proteins with phytochromes raises the possibility that light may regulate turnover of Aux/IAA proteins. To test whether light regulates SHY2/IAA3 protein turnover, we performed our turnover assay on extracts from dark-grown Arabidopsis seedlings treated with pulses of red light or shifted from darkness to light. We could not detect any difference in the rate of protein turnover (data not shown). In the TIR1::myc pull-down assay, SHY2/IAA3 protein interacted similarly with TIR1::myc protein expressed from either dark-grown or red light-treated seedlings (Figure 3d). Auxin still promoted this interaction in dark-grown seedlings as it did in light-grown seedlings (Figure 3d). Thus, we have detected no effect of light on auxin-regulated turnover of SHY2/IAA3.


Our data show that AXR1, TIR1, and the proteasome promote turnover of SHY2/IAA3 protein. In the pull-down assay, SHY2/IAA3 interacted with TIR1, the F-box protein component of the SCFTIR1 E3 ubiquitin ligase complex, and in the turnover assay, axr1 and tir1 mutations increased SHY2/IAA3 protein stability. The proteasome inhibitor MG132 also stabilized the protein. These results suggest that SCFTIR1-mediated ubiquitination targets SHY2/IAA3 to the 26S proteasome for degradation. The shy2-2 mutation decreased interaction with TIR1, indicating that domain II of SHY2/IAA3 destabilizes the protein by targeting it to SCFTIR1. This is consistent with the higher steady-state SHY2/IAA3 protein level previously observed in shy2-2 mutant plants (Colón-Carmona et al., 2000).

Auxin increased interaction of SHY2/IAA3 or shy2-2 with TIR1::myc. This auxin response was insensitive to the protein synthesis inhibitor cycloheximide, and also occurred in extracts, implying that auxin regulates the interaction post-translationally. Consistent with this idea, Gray et al. (2001) reported increased interactions between glutathione-S-transferase (GST)::AXR2/IAA7 or GST::AXR3/IAA17 fusion proteins and TIR1::myc within 5 min after auxin treament of seedlings. Together with work on AXR2/IAA7, AXR3/IAA17, and PS-IAA6::LUC fusion proteins (Dharmasiri et al., 2003; Gray et al., 2001; Zenser et al., 2001), our results suggest that SCFTIR1 targets multiple Aux/IAA proteins for auxin-induced ubiquitination and turnover.

The TIR1::myc pull-down assays were performed with transiently induced and presumably overexpressed TIR1::myc, and the SHY2/IAA3 protein also interacted with TIR1::myc carrying a mutation (P10A) in the F-box that decreases interaction of TIR1 with ASK1 or ASK2 (Gray et al., 1999). These results imply that free TIR1 can interact with potential substrates without incorporation into the SCFTIR1 complex. Such Ask/Cullin/Rbx1 complexes might then be free, together with other F-box proteins, to ubiquinate alternative substrates. Given the large number of F-box proteins encoded in the Arabidopsis genome (Gagne et al., 2002; Risseeuw et al., 2003), such a mechanism might be important for efficient protein turnover in vivo, and might also allow regulation of both the TIR1-Aux/IAA and TIR1–ASK interactions.

The turnover assay we have developed provides a novel tool to characterize regulation of Aux/IAA protein degradation. Its principal utility is to screen candidate mutations for effects on turnover, and the assay successfully revealed decreased SHY2/IAA3 turnover in axr1 and tir1 extracts. However, the assay did not reveal effects of auxin or juglone, or of the shy2-2 mutation. One possible explanation for this is that it relies on multiple biochemical steps, and the SHY2/IAA3–TIR1 interaction may not be rate-limiting in these cases. Nevertheless, it may be possible to modify this assay to improve its ability to detect additional components of SHY2/IAA3 turnover.

The TIR1::myc pull-down assay cannot be applied to multiple genotypes without first crossing the TIR1::myc transgene into various backgrounds, but does monitor a particular regulated step in the turnover pathway and readily revealed quantitative differences in interaction. Moreover, auxin regulation of the SHY2/IAA3–TIR1::myc interaction in cell extracts provided a means to dissect the regulation of this interaction biochemically. As an example of this, we have found that juglone very effectively inhibits the interaction. Juglone is produced by black walnut trees and inhibits growth of other plants nearby. It covalently attaches to sulfhydryl groups of cysteine residues in target proteins (Hennig et al., 1998). One known target of juglone is the parvulin class of peptidyl prolyl isomerases (Hennig et al., 1998; Metzner et al., 2001; Yao et al., 2001), and our results therefore raise the possibility that these enzymes might regulate SHY2/IAA3 protein turnover. This is an attractive hypothesis because two adjacent proline residues in conserved domain II (one of which is mutated in shy2-2) are important for the instability of Aux/IAA proteins (Ramos et al., 2001). One caveat to this model is that animal and plant parvulins act on phosphorylated substrates (at least for the peptide substrates tested) (Landrieu et al., 2000; Metzner et al., 2001; Yao et al., 2001), whereas we have found no evidence for a phosphorylation requirement for SHY2/IAA3 recognition of TIR1::myc, and furthermore domain II lacks an obvious phosphorylation site (Ramos et al., 2001). Juglone probably has additional targets that have not been identified (Chao et al., 2001). Juglone also inhibits the TIR1::myc interaction with AXR2/IAA7 (Dharmasiri et al., 2003), suggesting that whatever the relevant target of juglone may be, it regulates turnover of multiple Aux/IAA proteins.

The ability of phytochromes A and B to interact with SHY2/IAA3, the photomorphogenic phenotypes of shy2-1 and shy2-2 mutants, and regulation of SHY2/IAA3 expression by light all suggest that light regulates some aspect of SHY2/IAA3 protein function. However, our data revealed no light regulation of the interaction between SHY2/IAA3 and TIR1::myc proteins, or of SHY2/IAA3 protein turnover. Further work will be required to assess whether light regulates SHY2/IAA3 turnover or some other aspect of SHY2/IAA3 activity, such as nuclear localization or dimerization with ARF proteins.

Experimental procedures

Production of recombinant SHY2/IAA3 and shy2-2 mutant proteins and anti-SHY/IAA3 antibodies

SHY2/IAA3 cDNA fragment was amplified by PCR from a cDNA clone (Abel et al., 1995) using primers 5′-GCTCTAGAATGGATGAGTTTGTTAACC-3′ and 5′-TCGCCCGGGTACACCACAGCCTAAACC-3′ (underlined are XbaI and SmaI sites, respectively). shy2-2 cDNA fragment was obtained by RT-PCR from shy2-2 RNA using the same primers. These fragments were digested with XbaI and SmaI, and cloned in frame into pTYB2 vector with NheI (which creates a compatible cohesive end with XbaI) and SmaI sites in the IMPACT™-CN protein expression system (New England Biolabs, Beverly, MA, USA), so that the intein-CBD was at the C-terminus of the fusion proteins. These cloning steps introduced two extra amino acids (Ala and Arg) to the N-terminus of SHY2/IAA3 or shy2-2 proteins, and two extra amino acids (Pro and Gly) to the C-terminus of the proteins.

The plasmids were introduced into E. coli strain ER2566 (Koncz and Schell, 1986), and the E. coli cells were grown at 37°C in Luria-Bertani (LB) medium (1% tryptone, 0.5% yeast extracts, and 0.5% NaCl) supplemented with 100 µg ml−1 ampicillin until the OD600 of the culture reached 0.5–0.8. Fusion protein expression was then induced with 0.3 mm isopropyl beta-d-thiogalactopyranoside (IPTG) for 16 h at 15°C. The induced cells were collected by centrifugation at 5000 g for 10 min at 4°C. The pellets were re-suspended in 1/10 culture volume of lysis buffer (20 mm NaPO4 (pH 7.4), 0.5 m NaCl, 1 mm EDTA, and 0.2 mm phenylmethyl sulfonyl fluoride (PMSF), lysed in a French press, and centrifuged at 20 000 g for 30 min at 4°C. Supernatant from 1 l culture was passed over a column containing 10 ml chitin beads (New England Biolabs) at a flow rate of 0.5 ml min−1 at 4°C. After washing with 10 bed volumes of column buffer (20 mm NaPO4 (pH 7.4), 1 m NaCl, and 1 mm EDTA), the chitin columns were flushed with 2 bed volumes of cleavage buffer (20 mm NaPO4 (pH 7.4), 0.5 m NaCl, and 1 mm EDTA), and then incubated in 1 bed volume of cleavage buffer supplemented with 50 mm DTT (diluted in the cleavage buffer from 1 m DTT) at 4°C for 16 h. SHY2/IAA3 and shy2-2 proteins were eluted from the chitin columns with one bed volume of cleavage buffer without DTT and were dialyzed against the cleavage buffer without DTT overnight at 4°C. Two to four milligrams of SHY2/IAA3 protein was obtained per liter of induced culture.

Purified SHY2/IAA3 protein was used to immunize New Zealand White Rabbits by standard procedures by Covance (Denver, PA, USA). The crude antiserum was used at 1 : 8000 dilution for Western analysis. The serum was also purified by affinity chromatography using bacterially expressed SHY2/IAA3 protein immobilized on CNBr-activated Sepharose-4B (Amersham, Piscataway, NJ, USA; Harlow and Lane, 1988).

SHY2 and shy2-2 cDNA fragments were also amplified using primers 5′-GCTCTAGAATGGATGAGTTTGTTAACC-3′ and 5′-CCCTCGAGGGGTGTCTCTGTTAGATTTCTTG-3′ (underlined are XbaI and XhoI sites, respectively) and cloned in frame into pMAL-c2 vector (New England Biolabs) with XbaI and SalI (which creates a compatible cohesive end with XhoI) sites to produce MBP::SHY2/IAA3 fusion proteins. The plasmids were introduced into E. coli strain DH5α, and the cells were grown at 37°C in LB medium/0.2% glucose/100 µg ml−1 ampicillin until the OD600 of the culture reached 0.5–0.8. Fusion protein expression was then induced with 0.3 mm IPTG for 16 h at 23°C. The induced cells were collected by centrifugation at 4000 g for 20 min at 4°C. The pellets were re-suspended in 1/20 culture volume of buffer (20 mm Tris–HCl (pH 7.4), 0.2 m NaCl, 1 mm EDTA, and 0.2 mm PMSF), lysed in a French press, cleared by centrifugation at 20 000 g for 30 min at 4°C, and analyzed by SDS–PAGE and Western blots.

SDS–PAGE and immunoblot analysis

Proteins were resolved on SDS gels containing 12% acrylamide, and then either stained by Coomasie Brilliant Blue G-250 (Bio-Rad, Hercules, CA, USA) or by silver staining (Ausubel et al., 1987). For immunoblots, the proteins were transferred to BA-S83-supported nitrocellulose membranes (Schleicher and Schuell, Keene, NH, USA), and detected either with enhanced chemiluminescence (ECL Western blotting detection reagents, Amersham Pharmacia Biotech) or by using colorimetric reagents as described previously by Reed et al. (2000).

Crude AtCUL1 and ASK1 antisera were kindly provided by Mark Estelle (Indiana University) and Bill Crosby (Plant Biotechnology Institute, Saskatoon, Canada), and were used at 1 : 5000 for immunoblot analysis. Anti-c-myc monoclonal antibody 9E10 was purchased from the Tissue Culture Facility at University of North Carolina at Chapel Hill Lineberger Comprehensive Cancer Center and was used at a final concentration of 5 µg ml−1.

In vitro turnover assay

Nine-day-old seedlings grown in MS/sucrose liquid medium (Tian et al., 2002) were homogenized in ice-cold extraction buffer (50 mm Tris–HCl (pH 7.5), 150 mm NaCl, 0.5% NP-40, and 2 mm PMSF). The resulting homogenate was cleared by centrifugation at 14 000 g for 15 min at 4°C. Two hundred micrograms of crude extract was mixed with 0.3 µg purified SHY2/IAA3 protein in a total volume of 200 µl, and the mixture was incubated at 4°C with gentle agitation. At each time point, 20 µl was removed and re-suspended in 4 µl of 6× SDS–PAGE sample buffer, heated at 95°C for 5 min, and subjected to SDS–PAGE and immunoblotting.

Auxin or inhibitors were added either to seedlings grown in MS/sucrose liquid medium for 0.5–2 h at 23°C with moderate shaking before total protein was extracted, or directly to plant extracts, instead of the seedlings, for 0.5–2 h before performing the turnover assay. Concentrations used were 20 µm IAA, 100 µm MG132, 1 µm staurosporine, 50 µm cantharidin, 100 nm okadaic acid, and 10 mm NaF. Calf intestine alkaline phosphatase (5 units in 1× buffer #3, New England Biolabs) was added to the extracts. For MG132, treating the plant extracts inhibited degradation more effectively than treating seedlings before extraction, and therefore was presented in Figure 2(a).

Pull-down assays

Transgenic glucocorticoid-inducible tir1-1[TIR1::myc] seedlings were grown in MS/sucrose liquid medium for 7 days and induced with 0.03 mm dexamethasone for 24 h as described by Gray et al. (1999). Crude Arabidopsis extract was prepared as above for the turnover assay. One hundred micrograms of the extract (approximately 50 µl) was mixed with chitin beads carrying SHY2/IAA3::intein-CBD or shy2-2::intein-CBD protein (about 5 µl in elution buffer, corresponding to about 1.5 µg purified SHY2/IAA3 or shy2-2 protein) in a total volume of 100 µl extraction buffer supplemented with 10 µm MG132, 2 mm PMSF, and 5 µg ml−1 Leupeptin. In the experiments shown in Figure 3(a), the amounts of all components were scaled up 10-fold. The mixture was incubated for 2 h at 4°C and 0.5 h at room temperature with gentle shaking. The chitin beads were then washed six times with PBS buffer (137 mm NaCl, 3 mm KCl, 1.5 mm KH2PO4, 8 mm Na2HPO4·7H2O (pH 7.5)), and then incubated in PBS buffer containing 50 mm DTT overnight at 4°C to activate self-cleavage of the intein. The chitin beads were spun down by centrifugation, and the supernatant was subjected to SDS–PAGE and immunoblots.

For auxin or cycloheximide treatments, dexamethasone-treated seedlings were washed two to three times with MS/sucrose medium, and incubated in 20 ml of MS/sucrose medium with gentle shaking at 23°C in the presence of 20 µm IAA (from 20 mm stock in ethanol), 200 µg ml−1 cycloheximide (=0.7 mm, from 50 mm stock in ethanol), or both, for 2 h before total protein was extracted. These chemicals were added again directly to the pull-down assay. For auxin treatment in Figure 3(c), 2 µm or 20 µm IAA was added directly to the incubation mixture at the start of the pull-down assay without pre-treatment of the seedlings, and incubated together with plant extracts and SHY2/IAA3 or shy2-2 fusion proteins for 3 or 6 h at 4°C.

For curcumin (Sigma, St Louis, MO, USA; 50 mm stock in DMSO), juglone (Sigma #H4,700-3; 10 mm stock in EtOH), and 3,4-dehydro-l-proline (Sigma#D4893; 100 mm in H2O) treatment in Figure 4, the chemicals were added directly to the incubation mixture containing 1 mg plant protein extract and 40 µl SHY2::intein-CBD chitin beads (corresponding to 12 µg purified SHY2/IAA3 protein) without pre-treatment of the seedlings, and incubated for 2 h at 4°C and 0.5 h at room temperature.

For phyB pull-down assays, crude Arabidopsis extract from 10-day-old seedlings was prepared as above. 0.4 mg extract (about 200 µl) was pre-cleared with 35 µl chitin beads at 4°C for 30 min, and then mixed with chitin beads carrying SHY2::intein-CBD or shy2-2::intein-CBD protein (about 35 µl in elution buffer, corresponding to 4.5 µg purified SHY2/IAA3 or shy2-2 protein) in a total volume of 250 µl extraction buffer supplemented with 10 µm MG132, 2 mm PMSF, and 5 µg ml−1 Leupeptin. The mixture was incubated for 1.5–2 h at 4°C and 0.5 h at room temperature with gentle shaking. The chitin beads were then washed with alternate high- (Tris extraction buffer plus 850 mm NaCl and 1% Tween-20) and low-salt buffer (Tris extraction buffer) at 4°C with gentle shaking for 10 min each time and eight times in total. After the final wash, the chitin beads were re-suspended in 35 µl of 2× SDS–PAGE sample buffer, heated at 95°C for 5 min, and subjected to SDS–PAGE and immunoblotting. For Figure 5(b), half (1/2) or one-fourth (1/4) amount of the SHY2::CBD or shy2-2::CBD protein was used in the assay.


We thank J. Tepperman and P. H. Quail for seeds of overexpressed phyB lines. This work was supported by NIH grant GM52456 to J.W.R. Q.T. received partial support from a W. C. Coker fellowship and the Alma Holland Beers scholarship fund administered by the Department of Biology at UNC-CH.