The tir1 mutant, like ibr5 37, is less responsive to auxin in primary root elongation inhibition and lateral root formation assays 38. To examine the genetic interaction between ibr5 and tir1, we crossed tir1-1 to ibr5-1 and examined the phenotypes of the resulting double mutant. We found that tir1 ibr5 plants were shorter than either parent (Figure 1A). ibr5 cotyledon vascularization defects were sometimes mildly enhanced by tir1 (Figure 1B). In addition, the tir1 ibr5 double mutant displayed enhanced resistance to root elongation inhibition by 2,4-D and IBA (Figures 1C and 1D, Additional File 1), fewer lateral roots in response to IBA treatment (Figure 2A), and greater resistance to IBA inhibition of hypocotyl elongation in the dark (Figure 2B, Additional File 2) than either parent.

In addition to auxin resistance, ibr5 mutant roots are resistant to the phytohormone ABA 37. We found that tir1 also exhibited ABA resistance, and that tir1 ibr5 roots were more ABA resistant than either single mutant (Figure 2C, Additional File 2). Because ibr5-1 is likely to be a null allele 37, these results support a model in which IBR5 and TIR1 act separately to affect auxin and ABA responsiveness.

The axr1 mutant displays more extreme auxin-response defects than tir1 or ibr5, with restricted plant height, reduced apical dominance, dramatic vascularization defects, striking auxin resistance, and a longer root than wild type on unsupplemented media 29. To examine the genetic interaction between axr1 and ibr5, we crossed axr1-3 to ibr5-1. The double mutant had similar plant height to axr1-3 (Figure 3A), but leaf epinasty (data not shown) and cotyledon vascular defects (Figure 3B) were more extreme in axr1 ibr5 compared to either parent. Further, axr1 ibr5 had a longer root on unsupplemented media than either parent (Figure 3C), consistent with the possibility that resistance to endogenous auxin was enhanced. Resistance to the auxins 2,4-D and IBA was not obviously enhanced in the double mutant when considering the longer root on unsupplemented media (Figures 3C and 3D, Additional File 1). Moreover, axr1 ibr5 did not display enhanced resistance to IBA-induced lateral root formation (Figure 2A, Additional File 2), but did exhibit slightly enhanced resistance to the inhibition by IBA of hypocotyl elongation in the dark (Figure 2B, Additional File 2). Like ibr5, axr1 is resistant to ABA inhibition of root elongation 37, and axr1 ibr5 had similar ABA resistance as both parents (Figure 2C, Additional File 2).

The aux1 mutant displays marked resistance to the auxins that are brought into cells by the AUX1 transporter, such as 2,4-D and IAA 333439, but responds normally to 1-naphthaleneacetic acid (NAA), which is not transported by AUX1 333435. Although IBA does not appear to be an AUX1 substrate 3540, the aux1 mutant is moderately IBA resistant 36, probably because the IBA that enters cells is converted to IAA. aux1 mutant roots are agravitropic and longer than wild-type roots on unsupplemented media 39, but aux1 aerial parts resemble wild type. To examine the genetic interaction between aux1 and ibr5, we crossed ibr5-1 to aux1-7. Although aux1 plants attain normal height, adult aux1 ibr5 plants were shorter than either parent (Figure 4A). Moreover, aux1 ibr5 seedlings had longer roots on unsupplemented media than either parent (Figure 4C). Resistance to the auxins 2,4-D and IBA was not obviously enhanced in the double mutant when considering the longer root on unsupplemented media and the complete resistance of aux1 to the concentrations of 2,4-D tested (Figures 4C and 4D, Additional File 1). We examined lateral root production in aux1 ibr5 and found that aux1 did not markedly enhance ibr5 defects (Figure 2A, Additional File 2). Similarly, the ibr5 cotyledon vascular development defects did not appear to be enhanced by aux1 (Figure 4B). Hypocotyls of dark-grown aux1 responded like wild type to IBA, and aux1 ibr5 responses resembled those of ibr5 (Figure 2B, Additional File 2). aux1 was unresponsive to the ABA concentrations tested (Figure 2C, Additional File 2); thus we did not determine if ibr5 enhanced aux1 ABA resistance.

ibr5 seedlings grown on unsupplemented medium display reduced accumulation of DR5:GUS 37, a construct in which the GUS reporter is driven from a synthetic auxin-responsive promoter 13. We compared wild-type and ibr5 DR5:GUS auxin responses in roots of light-grown seedlings and hypocotyls of dark-grown seedlings. Two-hour treatments with various auxins increased DR5:GUS activity in wild-type roots (Figure 5A) and hypocotyls (Figure 5B). In contrast, ibr5 showed reduced induction of DR5:GUS activity following auxin treatment in both roots and hypocotyls (Figures 5A and 5B).

This reduced induction of DR5:GUS activity suggests that ibr5 misregulates at least some auxin-regulated transcripts. Thus, we examined basal levels and auxin responsiveness of endogenous IAA1, IAA2, and GH3.3 transcripts in wild-type and ibr5 seedlings. Although we detected subtle reductions in basal levels of IAA1 and IAA2 transcripts in ibr5 in several trials (data not shown), we did not detect dramatic differences in these experiments. In any case, IAA1, IAA2, and GH3.3 transcripts eventually reached similar maximal values in wild type and ibr5 (data not shown), suggesting that IBR5 is not required for full response to high auxin levels.

Gain-of-function mutations that stabilize any of several Aux/IAA proteins can confer dominant auxin resistance (reviewed in 18). Moreover, Aux/IAA repressor proteins or Aux/IAA-reporter fusion proteins are stabilized in numerous other auxin-resistant mutants, including tir1 16, axr1 16, ecr1 28, afb1, afb2, and afb3 21, cul1 41, eta2/cand1 42, eta3/sgt1b 43, and aar1 44.

Because auxin-responsive transcripts are reduced in ibr5, we sought to analyze Aux/IAA stability in the ibr5 mutant. We crossed ibr5-1 to a line expressing the AXR3/IAA17 Aux/IAA protein N-terminal degron region fused to β-glucuronidase driven by a soybean heat-shock promoter (HS:AXR3NT-GUS; 16). Eight-day-old seedlings were heat shocked to induce reporter transcription, and then either mock treated or auxin treated. In seedlings with intact auxin signaling, the auxin-induced disappearance of AXR3NT-GUS activity reflects the targeting of the reporter to the 26S proteasome for degradation 16. Mutants with auxin-response defects, such as tir1 and axr1, show increased reporter activity after induction and reduced destabilization of AXR3NT-GUS upon auxin treatment 16, consistent with the axr1 defect in accumulating normal levels of auxin-responsive transcripts following auxin treatment 4546. Intriguingly, unlike in previously characterized auxin-response mutants, we did not detect increased AXR3NT-GUS activity in ibr5. In fact, AXR3NT-GUS appeared to be less active in ibr5 than in wild-type roots 10 or 20 minutes following heat shock (Figure 6A). This decrease in AXR3NT-GUS activity was also apparent in tir1 ibr5, suggesting that ibr5, although enhancing tir1 auxin resistance in root elongation (Figures 1C, D, and 6B), suppressed tir1 AXR3NT-GUS accumulation (Figure 6A). Response to 2,4-D for each of these lines was as expected, with tir1 ibr5 (HS:AXR3NT-GUS) showing enhanced resistance compared to the intermediate resistance of either parent (Figure 6B).

The reduced AXR3NT-GUS activity in ibr5 was apparent immediately following the 2-hour heat shock used to induce reporter expression (data not shown). Because a heat-responsive promoter drives the AXR3NT-GUS construct, we tested whether the lack of AXR3NT-GUS activity in ibr5 could be explained by a reduced transcriptional response to heat in the mutant. We introduced a construct altered to contain the stabilizing axr3-1 mutation (HS:axr3-1NT-GUS; 16) into ibr5-1 by crossing. The proline to leucine substitution in axr3-1 47 confers reporter stability by decreasing axr3-1NT-GUS interaction with SCF^TIR1 16. We found similar axr3-1NT-GUS activity in wild type and ibr5 with or without auxin treatment (Figure 6C), suggesting that the transgenes were efficiently transcribed in response to the heat stimulus, and that the decreased AXR3NT-GUS activity in ibr5 was not caused by reduced transcription following heat shock. Moreover, treatment with the proteasome inhibitor MG132 restored AXR3NT-GUS activity in ibr5 and tir1 ibr5 to near wild-type levels (Figure 6D), again suggesting that the ibr5 defects in AXR3NT-GUS activity were not due to reduced transgene transcription.

IAA28 was originally identified because the iaa28-1 gain-of-function mutation confers auxin resistance and impedes lateral root production 48. Like typical Aux/IAA proteins, IAA28 has a transcriptional repressor domain 15 and can confer auxin-enhanced instability to a luciferase reporter 49. To test whether the lack of AXR3NT-GUS stabilization in ibr5 was accompanied by stability effects on other Aux/IAA proteins, we crossed ibr5-1, tir1-1, and axr1-3 to a wild-type line carrying a c-Myc epitope-tagged version of IAA28 driven from IAA28 regulatory sequences (Figure 7A) and isolated homozygous mutants carrying the reporter transgene. These lines responded to 2,4-D as expected (Figure 7B). Because IAA28 is primarily expressed in roots 48, we examined accumulation of IAA28myc in wild-type and mutant roots following mock- or auxin-treatment of 10-day-old seedlings. As expected, we found that IAA28myc disappeared rapidly following auxin treatment in wild type (Figure 7C). We found lower IAA28myc levels in ibr5 than wild type in the absence of added auxin; this protein disappeared rapidly upon auxin treatment and was not detected after 10 minutes of treatment (Figure 7C). Although we had expected IAA28myc to be stabilized in tir1 and axr1, we found IAA28myc levels similar to wild-type levels that decreased in response to auxin treatment (Figure 7C) in both mutants, consistent with the observation that neither of these mutants is completely insensitive to auxin (e.g., Figure 7B). We examined IAA28 mRNA levels in these lines and found reduced IAA28 transcript levels in axr1 and, to a lesser extent, in tir1 (Figure 7D). Transcript levels were not perfectly correlated with IAA28 protein levels, suggesting differences in protein stability in these mutants. Treatment with MG132 increased IAA28myc protein levels in both wild type and ibr5 (Figure 7E), suggesting that IAA28myc is degraded via the 26S proteasome in wild type and consistent with the possibility that the proteasome contributes to the reduced IAA28myc levels in ibr5. However, the inability of MG132 to fully restore IAA28myc to wild-type levels in ibr5 (Figure 7E) suggests that the reduced IAA28 mRNA level in ibr5 (Figure 7D) contributes to the reduced IAA28myc accumulation in this mutant (Figure 7C). The striking lack of AXR3NT-GUS and IAA28myc stabilization in ibr5 is consistent with the possibility that auxin-regulated transcription is reduced in ibr5 via a mechanism that does not involve Aux/IAA protein stabilization.

Dual specificity protein phosphatase (DSP) proteins dephosphorylate both threonine and tyrosine residues of phosphorylated proteins, often thereby inactivating them (reviewed in 50). DSP proteins contain a conserved aspartate residue and a separate, highly conserved signature motif of VxVHCx₂GxSRSx₅AYLM, with the cysteine and arginine residues participating with the conserved aspartate in catalysis. The cysteine of this signature begins the dephosphorylation process with a nucleophilic attack on the phosphorus atom of the phosphotyrosine or phosphothreonine substrate. Thus, disruption of this conserved cysteine results in catalytic inactivity in many DSP proteins (reviewed in 51), including the IBR5 relative, DsPTP1 52.

The DSP active site motif VxVHCx₂GxSRSx₅AYLM is present in IBR5 (Figure 8A), allowing us to identify the presumptive active site cysteine (C129) in IBR5. To test whether IBR5 phosphatase activity is required for normal auxin responses in vivo, we generated transgenic Wt and ibr5-1 lines expressing a Cys129 → Ser129 (C129S) substitution variant of IBR5 (IBR5^C129S) under the control of the strong 35S viral promoter. We anticipated that 35S:IBR5^C129S would not rescue ibr5-1 defects if IBR5 phosphatase activity were required to promote auxin responsiveness. We assayed ibr5-1 (35S:IBR5^C129S) lines expressing low (line A) and high (line B) IBR5^C129S levels (Figure 10C) for mutant phenotype rescue. Wild-type and ibr5-1 lines overexpressing unmodified IBR5 (35S:IBR5; 37) were used for comparison.

We found that IBR5 or IBR5^C129S overexpression did not noticeably alter wild-type adult morphology (Figure 8B). Moreover, IBR5 or IBR5^C129S overexpression in wild type did not alter any of the hormone-response phenotypes examined (Figure 8D, F). As expected, we found that ibr5-1 (35S:IBR5) plants were restored to wild-type height (Figure 8B) and had normal root responses to auxin and ABA (Figure 8E, F). ibr5-1 (35S:IBR5^C129S) line A (low expressor) exhibited restored plant height but had similar leaf epinasty to ibr5-1, whereas ibr5-1 (35S:IBR5^C129S) line B (high expressor) displayed rescue of ibr5-1 plant height (Figure 8B) and partial rescue of leaf epinasty (data not shown). Resistance to the inhibitory effects of 2,4-D and ABA on root elongation was not rescued in ibr5-1 (35S:IBR5^C129S) line A and only partially restored in ibr5-1 (35S:IBR5^C129S) line B (Figures 8E, F). The lack of full ibr5-1 rescue by IBR5^C129S is consistent with the possibility that IBR5 phosphatase activity is required for full auxin and ABA responsiveness. The partial ibr5-1 rescue observed when IBR5^C129S accumulates to high levels suggests that certain IBR5 functions do not require phosphatase activity. For example, IBR5^C129S may bind and sequester its substrate(s), thereby dampening normal substrate activity.

Arabidopsis thaliana accession Colombia (Col-0) was the wild type used for all experiments. The ibr5-1 mutant contains a nonsense mutation at IBR5 amino acid 42, resulting in a truncated product lacking the catalytic domain 37. The aux1-7 mutant contains a missense mutation resulting in glycine 459 being replaced by aspartic acid 32. The tir1-1 mutant contains a missense mutation resulting in glycine 147 being replaced by an aspartic acid 38. The axr1-3 mutant contains a missense mutation resulting in cysteine 154 being replaced by a tyrosine 26.

Surface-sterilized 68 seeds were plated on PNS (plant nutrient medium with 0.5% [w/v] sucrose) 69 solidified with 0.6% (w/v) agar. Hormones used were from 0.1-, 1.0-, or 100-mM stocks in ethanol, with ethanol-supplemented media used as controls, and all treatments normalized to the same ethanol content (less than 0.1 μL ethanol/mL medium). Seedlings were grown at 22°C under continuous light. Unless indicated otherwise, plates were incubated under yellow long-pass filters to slow the breakdown of indolic compounds 70. Plants were grown in soil (Metromix 200; Scotts, Marysville, OH) at 22 to 25°C under continuous illumination by cool-white fluorescent bulbs (Sylvania, Danvers, MA).

The pKSIBR5c construct 37 was mutated using oligonucleotide-directed mutagenesis 71 to alter the presumptive catalytic cysteine at amino acid position 129 to a serine using a primer 5'-CTTTCCCAGACAT

CG

The IAA28:IAA28myc construct 73 contains the IAA28 genomic region, including 3.1 kb of DNA 5' of the IAA28 coding sequence, subcloned into the pBIN19 plant transformation vector. The last exon of IAA28 has been modified in this construct to encode five copies of the c-Myc epitope immediately upstream of the IAA28 termination codon. The IAA28:IAA28myc plasmid was electroporated 71 into Agrobacterium tumefaciens GV3101, which was used to transform wild-type Col-0. Transformed seedlings were identified on PN medium supplemented with 12 μg/ml kanamycin after growth under white light. Homozygous single-insert lines were identified by examining the pattern of kanamycin resistance in subsequent generations. ibr5-1, tir1-1, and axr1-3 were crossed to a wild-type (Col-0) single-insert line to generate ibr5, tir1, and axr1 carrying the IAA28:IAA28myc construct.

All assays were conducted at least three times with similar results. For auxin-response root-elongation assays, seedlings were grown for 8 or 9 days on PNS with the indicated auxin concentrations and removed from the agar, and the lengths of the primary roots were measured. For ABA response root-elongation assays, seedlings were grown for 4 days on PNS to allow efficient germination, then were transferred to PNS supplemented with either ethanol or ABA. After an additional 4 days of growth, primary root lengths were measured.

In lateral root assays, seedlings were grown for 4 days on PNS, transferred to PNS supplemented with either ethanol or 10 μM IBA, and grown for an additional 4 days, after which lateral roots were examined under a dissecting microscope. Primordia emerged from the primary root were counted as lateral roots.

For hypocotyl elongation assays, seeds were plated on media supplemented with either ethanol or 20 μM IBA. After 1 day in the light, plates were wrapped with aluminum foil and incubated for an additional 4 days in the dark, after which seedlings were removed from the agar and hypocotyls lengths were measured.

To examine cotyledon vascular patterns, seedlings were grown for 8 days on PNS. Chlorophyll was removed using an ethanol series, and seedlings were cleared by incubating for one week at room temperature in chloral hydrate solution (80 g chloral hydrate, 20 mL glycerol, and 10 mL water). Cleared seedlings were mounted and photographed through a dissecting microscope.

The ibr5-1 mutant was crossed to aux1-7 39, axr1-3 30, and tir1-1 38, all in the Col-0 accession. Double mutants were identified by PCR analysis of F2 plants. Amplification of AUX1 with AUX1-3 (5'-CATGGGTCAACAAAGCTTTGGATTTTGTCC-3') and AUX1-4 (5'-TTCGTGACTTTTACTCCCTTCACGTATACG-3') yields a 464-bp product with two DpnII restriction sites in wild type and three in aux1-7. Amplification of AXR1 with the derived cleaved amplified polymorphic sequence primer 7475 AXR1-Acc1 (5'-AAACCAACTTAACGTTTGCAT

G

Surface-sterilized 68 seeds were plated on filter paper-lined PNS and grown under continuous illumination. After 6 days, the filter paper with seedlings was lifted off the agar surface. Seedlings were floated in liquid PN supplemented with mock (ethanol) or 200 nM IAA for 10 minutes. Seedlings were collected and ground with a mortar and pestle in liquid nitrogen. RNA was extracted using TriReagent (Sigma) according to the manufacturer's instructions. Total RNA (10 μg) was electrophoresed on a 1% agarose gel containing 0.37 M formaldehyde 71 and transferred to a positively-charged nylon 66 membrane (Roche Applied Science, Indianapolis, IN) in 20× SSC using capillary action. DIG-labeled probe was hybridized in DIG Easy Hyb buffer (Roche) overnight at 50°C and washed at moderate stringency according to the manufacturer's instructions. Probes were detected using an alkaline phosphatase-conjugated anti-DIG antibody (Roche) diluted 1:20,000 in Blocking Buffer (Roche), then visualized using a 1:200 dilution of CDP-Star (Roche).

DIG-labeled IAA28 probe was synthesized by PCR amplification using a PCR DIG Probe Synthesis Kit (Roche) according to the manufacturer's instructions from a cDNA template 48 using T1N24-20 (5'-CCATCGAACTGATGATTTTGGCC-3') and T1N24-21 (5'-CCTCCTTGTCACCAATTCACTTCC-3'), yielding a 525-bp product.

To visualize IBR5, protein was extracted from entire 2-day-old seedlings grown in 0.1% agar under white light by grinding frozen tissue with a pestle and adding one volume NuPAGE 2X LDS buffer (Invitrogen, Carlsbad, CA). Debris was pelleted by centrifugation for 4 minutes. The supernatant was heated to 100°C for 5 minutes. Protein extracts were separated by SDS-polyacrylamide gel electrophoresis beside Cruz markers (Santa Cruz Biotechnology, Santa Cruz, CA) using a NuPAGE 10% Bis-Tris gel and MES running buffer (Invitrogen).

Protein was transferred for 35 min at 24 V to a Hybond ECL nitrocellulose membrane (Amersham Pharmacia Biotech, Piscataway, NJ) using NuPAGE transfer buffer (Invitrogen). After blocking for 1 hour in 8% powdered milk in Tween Tris-buffered saline (TTBS; 71), the membrane was incubated overnight at 4°C with affinity purified IBR5 antibody 37 diluted 1:20 in blocking buffer. The membrane was then washed three times with TTBS and incubated with HRP-conjugated goat anti-rabbit secondary antibody (Santa Cruz Biotechnology) diluted 1:2000 for 1 hour, washed again as described before, then visualized using LumiGLO reagent (Cell Signaling, Beverly, MA). Membranes subsequently were incubated with an antibody against spinach (Spinacia oleracea) HSC70 (Stressgen Bioreagents SPA-817) diluted 1:5000, followed by HRP-conjugated goat anti-mouse secondary antibody (Santa Cruz Biotechnology) diluted 1:2000, and visualized using LumiGLO reagent (Cell Signaling).

To visualize IAA28myc, 10-day-old light-grown seedlings were removed from PNS medium and floated in liquid PN supplemented with either ethanol (mock) or 200 nM IAA. At the indicated time points, roots were excised and protein was extracted. For MG132 treatment, 3-day-old light-grown seedlings were removed from PNS medium and suspended in water containing either DMSO (mock) or 300 μM MG132. After 1 or 2 hours, protein was extracted from whole seedlings. In both experiments, protein was separated by SDS-polyacrylamide gel electrophoresis and transferred to a membrane as described above. After blocking for 1 hour in 8% powdered milk in TTBS, the membrane was incubated overnight at 4°C with monoclonal 9E10 anti-c-Myc antibody (Santa Cruz Biotechnology SC-40) diluted 1:500 and anti-HSC70 antibody (Stressgen Bioreagents SPA-817) diluted 1:250,000 in blocking buffer. The membrane was then washed three times with TTBS and incubated with HRP-conjugated goat anti-mouse secondary antibody (Santa Cruz Biotechnology) diluted 1:2000 for 4 hours, washed again as described before, then visualized using LumiGLO reagent (Cell Signaling).

Wild-type (Col-0) lines carrying HS:AXR3NT-GUS and HS:axr3-1NT-GUS and axr1-3 carrying HS:AXR3NT-GUS were described previously 16. To obtain additional lines carrying these reporters, ibr5-1 was crossed to wild type carrying HS:AXR3NT-GUS and to wild type carrying HS:axr3-1NT-GUS. tir1 ibr5 was crossed to wild type carrying HS:AXR3NT-GUS to obtain tir1-1 HS:AXR3NT-GUS and tir1 ibr5 HS:AXR3NT-GUS. Mutants were identified by PCR analysis of the F2 plants, as described above. The presence of HS:AXR3NT-GUS or HS:axr3-1NT-GUS was assayed by resistance to kanamycin and confirmed by GUS staining.

For histochemical assays, 8-day-old light-grown seedlings were removed from PNS plates and floated in 0.5 mL liquid PN medium diluted six-fold with water (for hormone response assays) or sterile water (for MG132 assays) contained in a 12-well plate. Two milliliters of prewarmed (37°C) one-sixth liquid PN or prewarmed water was added to each well, and the plates were incubated at 37°C for 2 hours. For the hormone response assays, seedlings were transferred to room temperature for 20 minutes prior to 10- and 20-minute mock (ethanol) or 100 nM IAA treatments, or 40-minute mock (ethanol) or 10 μM IBA treatment. For MG132 assays, seedlings were incubated in the dark, and either DMSO (mock) or MG132 (to 50 μM) was added to each well 1 hour into the heat treatment. After heat treatment, plates were transferred to room temperature for 2 hours. Seedlings were stained for GUS activity as previously described 76.