The main oxidative inactivation pathway of the plant hormone auxin

Inactivation of the phytohormone auxin plays important roles in plant development, and several enzymes have been implicated in auxin inactivation. In this study, we show that the predominant natural auxin, indole-3-acetic acid (IAA), is mainly inactivated via the GH3-ILR1-DAO pathway. IAA is first converted to IAA-amino acid conjugates by GH3 IAA-amidosynthetases. The IAA-amino acid conjugates IAA-aspartate (IAA-Asp) and IAA-glutamate (IAA-Glu) are storage forms of IAA and can be converted back to IAA by ILR1/ILL amidohydrolases. We further show that DAO1 dioxygenase irreversibly oxidizes IAA-Asp and IAA-Glu into 2-oxindole-3-acetic acid-aspartate (oxIAA-Asp) and oxIAA-Glu, which are subsequently hydrolyzed by ILR1 to release inactive oxIAA. This work established a complete pathway for the oxidative inactivation of auxin and defines the roles played by auxin homeostasis in plant development.

(a) HPLC chromatogram of the reaction products of the AtDAO1 enzyme. IAA-Asp was used as a substrate. oxIAA-Asp was detected as a reaction product. (b) SDS-PAGE analysis of purified AtDAO1 protein (34.6 kDa) expressed in E. coli harboring the pCold-AtDAO1 vector. This experiment was repeated three times independently with similar results. (c) Lineweaver-Burk plot and kinetics parameters of AtDAO1 for IAA-Asp. (n=3). Error bars, SE. (d) HPLC chromatogram of the reaction products of the AtDAO1 enzyme. IAA-Glu, IAA-Ala, and IAA-Leu were used as substrates. oxIAA-Glu, oxIAA-Ala, and oxIAA-Leu were detected as reaction products from IAA-Glu, IAA-Ala, and IAA-Leu, respectively. (n=3) (e) HPLC chromatogram of the reaction mixtures of AtDAO1. IAA-β-D-glucoside, IAA-Asp dimethyl ester (IAA-Asp-DM), and IAA-Asp were used as substrates. IAA-Asp was completely consumed after 15 min of reaction. A total of 98% IAA-β-D-glucoside and 97% IAA-Asp-DM remained after the AtDAO1 reaction for 30 min. (f) AtDAO1 oxidizes IAA to produce oxIAA as a reaction product. The oxIAA production rate by AtDAO1 was 5.51 ± 0.13 pmol/min/mg under our assay conditions. (n=4) (g) oxIAA-amino acids were detected as a diastereomeric mixture from the reaction mixture. One of the diastereomers of oxIAA-Asp was readily epimerized during sample preparation. The molecules are shown as space-filling models (c and d) or stick models (e-h). The IAA moiety was located at the same position among the IAA and IAA-amino acid conjugates. The 2-position of the IAA moiety was oriented to 2-OG and the magnesium ion (arrows). IAA could not fill the binding site. On the other hand, the amino acid moieties of IAA-Asp and IAA-Glu occupied the entire binding pocket of AtDAO1. Under this calculation condition, the affinity scores (Kd) of IAA, IAA-Asp, IAA-Glu, IAA-Leu, and IAA-Ala were estimated to be -6.1, -7.7, -7.1, -7.0, and -7.5, respectively. were less active in the root auxin response than diesters of IAA-Asp. WT and dao1-1 mutant seedlings were grown for 6 days on GM plates containing 20 μM compounds (for wild-type plants) or 2 μM compounds (for dao1-1 mutants). The different letters represent statistical significance at P<0.01 (Tukey's HSD test, n = 18). Scale bar, 10 mm. (c) Auxin-inducible DR5::GUS transgene expression in dao1dao2 and WT seedlings. Six-day-old seedlings were cultured for 8 h in liquid GM media with chemicals (μM). The auxin-induced GUS activity was visualized by X-Gluc. Scale bar, 1 mm. (d) The auxin-resistant mutants axr1-3, tir1 afb2, and slr1/iaa14 were resistant to IAA-Asp-DM. The WT and mutant plants were grown on GM media (4 g/L agar) containing IAA containing IAA-Asp-DM for 6 days. Relative root length is shown as the percentage of that in mock-treated plants (100%). The different letters represent statistical significance at P<0.005 (Tukey's HSD test, n = 15).
Supplementary Fig. 8. IAA-amino acid diesters showed auxin activity in Brachypodium distachyon and rice (Oryza sativa). (a) Brachypodium distachyon Bd21 seeds were cultured for 8 days on GM agar medium with hormones. Scale bar, 10 mm. (b) Oryza sativa was cultured for 5 days in distilled water containing hormones. The different letters represent statistical significance at P<0.001 (Tukey's HSD test), and n indicates the number of plants analyzed. Scale bar, 10 mm. Similar to IAA, both IAA-Asp-DM and IAA-Glu-DM inhibited root growth in B. distachyon and rice seedlings. Supplementary Fig. 9. AtDAO1 modulates auxin activity derived from IAA-Asp and IAA-Glu. (a) Lateral root formation in dao1-1 and AtDAO1-overexpressing plants (35S::AtDAO1). Five-day-old plants were incubated for an additional 3 days on a horizontal GM plate containing IAA and conjugates. IAA induced lateral root formation of WT (Col), dao1-1, and 35S::AtDAO1 plants to the same extent. IAA-Asp-DM and IAA-Glu-DM induced lateral root formation in WT (Col) and dao1-1 but not in 35S::AtDAO1 plants. The dao1-1 mutant was more sensitive to the conjugates than WT plants. The different letters represent statistical significance at P<0.001 (Tukey's HSD test, n = 11-14). Scale bar, 1 mm. (b) Phenotype of the AtDAO1 overexpression line. The seedlings were grown vertically for 7 days on GM plates containing 2 μM estradiol (ER) with or without IAA-Asp-DM (3 μM). ER was added to induce the AtDAO1 transgene in pMDC7::AtDAO1 in the dao1-1 mutant. Scale bar, 10 mm. (c, d) The primary root length of WT and 35S::AtDAO1 and dao1-1 mutants cultured for 6 days on GM plates (4 g/L agar) containing IAA-Asp-DM (c) and IAA-Glu-DM (d). Relative root length is shown as the percentage of that in mock-treated plants (100%). The different letters represent statistical significance at P<0.001 (Tukey's HSD test, n = 34-40). (e) 35S::AtDAO1 and WT roots showed the same sensitivity to IAA. Seedlings were cultured on a vertical GM plate (6 g/L gellan gum) containing IAA for 6 days. The different letters represent statistical significance at P<0.001 (Tukey's HSD test, n = 22-28). AtDAO1 predominantly inactivates IAA-amino acid conjugates rather than IAA in planta. The values shown are the means ± SD (***P < 0.001, ns, not significant, two-tailed Student's ttest, n = 3).

Supplementary Fig. 13. Endogenous amounts of IAA metabolites in rice (Oryza sativa) and Brachypodium distachyon.
Our working model of the IAA inactivation pathway is shown. (a-f) Oryza sativa cv. Nipponbare was grown on agar (3 g/L) for 5 days at 25 °C under continuous light after 2 days of incubation at 28 °C. The plants were incubated in water with 1 μM compounds for 24 h. After extensive washing of seedlings with fresh water, the root was excised. The metabolites in the root were analyzed with LC-MS/MS. The values shown are the means ± SD (***P < 0.001, **P < 0.01, ns: not significant, Two-tailed Student's t-test [vs mock treatment], n = 4). (g-l) Brachypodium distachyon Bd21 was grown hydroponically with 1/10 GM medium (without sucrose) for 15 days at 24 °C under continuous light. The plants were incubated in water with 1 μM compounds for 24 h. After extensive washing of seedlings with fresh water, the root was excised. The metabolites in the root were analyzed with LC-MS/MS. The values shown are the means ± SD (***P < 0.001, **P < 0.01, ns: not significant, Two-tailed Student's t-test [vs mock treatment], n = 3). (a) Effects of IAA-Glu-DM on root growth of dao1 ilr/ill mutants (dao1-1, dao1 dao2, iar3 dao1-1, ill2 dao1-1, and ilr1 dao1-1). The seedlings were grown for 6 days on GM plates (4 g/L agar) containing IAA-Glu-DM. Relative root length is shown as the percentage of that in mock-treated plants (100%). The ilr1 mutation decreased high sensitivity of root inhibition in the dao1-1 mutant. The different letters represent statistical significance at P<0.01 (Tukey's HSD test, n = 20). (b) Effects of IAA-Asp-DM and IAA-Glu-DM on lateral root formation of dao1 ilr/ill mutants (dao1-1, ilr1 iar3 dao1-1, and ilr1 ill2 iar3 dao1-1). The seedlings were grown vertically for 6 days on GM plates and then transferred to a horizontal GM plate containing compounds. The seedlings were cultured for another 3 days. The root length and lateral root number were measured. The different letters represent statistical significance at P<0.01 (Tukey's HSD test, n =20). (c and d) Overexpression of the ILR1-GFP fusion protein enhanced the sensitivity of the dao1 dao2 mutant to IAA-Asp and IAA-Glu and their diesters (IAA-Asp-DM and IAA-Glu-DM). The seedlings were grown vertically for 8 days on GM plates (6 g/L gellan gum) containing IAA conjugates. Primary root length was measured. Scale bar, 10 mm. The different letters represent statistical significance at P<0.001 (Tukey's HSD test, n = 19-20).
Supplementary Fig. 18. The ilr1 iar3 loss-of-function mutation restored dao1-1 phenotypes. (a) Root hair length of wild-type (Col), dao1-1, and ilr1 iar3 dao1-1 triple mutant plants. Scale bar, 1 mm. The seedlings were grown vertically for 9 days on GM plates (6 g/L gellan gum). Primary root hair length was measured. The different letters represent statistical significance at P<0.001 (Tukey's HSD test). The values shown are the means, and n indicates the number of root hairs analyzed. Scale bar, 1 mm. (b) Lateral root density in wild-type, dao1-1, and ilr1 iar3 dao1-1 mutant plants. The seedlings were grown vertically for 9 days on GM plates. The number of lateral roots and root length were measured. Scale bar, 10 mm. The different letters represent statistical significance at P<0.01 (Tukey's HSD test). P=0.348 for letter a. (c) Phenotypes of primary inflorescences from 40-day-old WT (Col), dao1-1, ilr1 dao1-1, ilr1 iar3 dao1-1 and ilr1 ill2 iar3 dao1-1 mutant plants (Col background). The ilr1 mutation restored less fertility in primary inflorescences in the dao1-1 mutant. At least five plants from each line were examined. Scale bar, 10 mm. This experiment was repeated three times independently. Supplementary Fig. 19. GST-ILR1 hydrolyzed both oxIAA-amino acid conjugates and IAA-amino acid conjugates.

Synthesis of chemicals
General experimental conditions. 1 H and 13 C-NMR spectra were recorded on ECS400 and ECZ400 spectrometers (JEOL, Japan). Chemical shifts are shown as δ values from TMS as the internal reference. Peak multiplicities are quoted in Hz. Mass spectra were measured on autoflex speed MALDI-TOF MS (Bruker, Japan), and Agilent 6420 Triple Quad LC-MS (Agilent Technologies, USA). Column chromatography was carried out on columns of silica gel 60 (230-400 mesh, Merck, Japan). All chemicals were purchased from Tokyo Chemical Industry Japan (Japan), FUJIFILM Wako Pure Chemical (Japan) and Sigma-Aldrich Japan (Japan) unless otherwise stated.  was synthesized according to the method previously described 5 . The 4-(1,3-dioxoisoindolin-2-yl)butanoic acid was obtained as colorless powder (1358 mg, 75% yield). 1  Kakeimide was found to be potent inhibitor specific to GH3 enzyme that catalyzed the conjugation of IAA with an amino acid. Kakeimde inhibited GH3 enzymatic activity in competition with IAA. Recombinant GH3.6 protein was purified by TALON metal affinity resin from the culture lysate of E. coli BL21 harboring pCold-GH3.6. The enzyme reaction was performed in the following assay condition; GH3.6 protein: 1.5 μg/mL, 50 mM Tris-HCl (pH 8.6), 3 mM MgCl 2 , 3 mM ATP, various concentration of IAA, 3 mM L-aspartate, 1 mM dithiothreitol, 30 °C, 30 min 6 . The product, IAA-Asp was analyzed with HPLC system (EXTREMA, JASCO Japan). The Ki values for AtGH3.6 (Km=55.8 μM for IAA) was determined to be 48 nM (Supplementary table 3

oxIAA-L-Glu
oxIAA-L-Glu was synthesized with the same procedure for oxIAA-L-Ala. oxIAA-L-Glu was obtained as a diastereomeric mixture (1 : 1) of oxIAA-L-Glu at C-3 position (78 mg, 54% yield) as a pale yellow solid. 1   Red colored sequence shows the attB1 and attB2 adapter sequence for adaptor PCR (2 rounds of PCR with shorter oligos) using attB adapter primers.