A noncanonical auxin-sensing mechanism is required for organ morphogenesis in Arabidopsis

Abstract

Tissue patterning in multicellular organisms is the output of precise spatio-temporal regulation of gene expression coupled with changes in hormone dynamics. In plants, the hormone auxin regulates growth and development at every stage of a plant's life cycle. Auxin signaling occurs through binding of the auxin molecule to a TIR1/AFB F-box ubiquitin ligase, allowing interaction with Aux/IAA transcriptional repressor proteins. These are subsequently ubiquitinated and degraded via the 26S proteasome, leading to derepression of auxin response factors (ARFs). How auxin is able to elicit such a diverse range of developmental responses through a single signaling module has not yet been resolved. Here we present an alternative auxin-sensing mechanism in which the ARF ARF3/ETTIN controls gene expression through interactions with process-specific transcription factors. This noncanonical hormone-sensing mechanism exhibits strong preference for the naturally occurring auxin indole 3-acetic acid (IAA) and is important for coordinating growth and patterning in diverse developmental contexts such as gynoecium morphogenesis, lateral root emergence, ovule development, and primary branch formation. Disrupting this IAA-sensing ability induces morphological aberrations with consequences for plant fitness. Therefore, our findings introduce a novel transcription factor-based mechanism of hormone perception in plants.

Keywords: Arabidopsis; ETTIN; IAA; auxin signaling; plant development; transcription factor complex.

Figure 1.

ETT and IND genetically interact to regulate gynoecium development. (A–D) Scanning electron micrographs (SEMs) of apices from stage 12 gynoecia of Col-0 (A), ett-3 (B), ind-2 (C), and ett-3 ind-2 (D). (E–H) Confocal images of stage 11 gynoecium (E) from double-transgenic lines expressing pETT::ETT-CFP (F) and pIND::IND-YFP (G) and their colocalization in the nucleus (H). Insets show close-ups of nuclei. (stg) Stigma; (sty) style; (va) valve. Bars, 100 µm.

Figure 2.

ETT and IND function together to regulate target genes during gynoecium development. (A) Phylogenetic shadowing using mVISTA of a 2-kb genomic region upstream of the translational start site of the PID gene with pairwise alignments of Arabidopsis thaliana with Arabidopsis lyrata, Capsella rubella, Camelina sativa, and Brassica rapa. Regions 1 and 2 (R1 and R2) are indicated by shaded areas. (B) ChIP with the pETT::ETT-GFP line showing enrichment of a fragment containing the conserved AuxRE sites at −429 and −447. The WUS promoter was used as a positive control. Error bars show standard deviation. (*) P < 0.01; (**) P < 0.001. (C–F) In situ hybridization of PID mRNA at the apex of stage 8 gynoecia from Col-0 (C), ett-3 (D), ind-2 (E), and ett-3 ind-2 (F). (se) Sepal; (st) stamen; (stg) stigma; (sty) style; (va) valve. Bars, 50 μm.

Figure 3.

ETT and IND proteins interact. (A–C) BiFC with pYFPN-ETT and pYFPC-IND (A) and respective negative controls pYFPN-ETT + pYPFC empty (B) and pYFPN empty + pYFPC-IND (C). (D, left) Y2H assay showing positive interaction between IND-AD and ETT-BD and between IND-AD and SPT-BD. (Right) Negative controls (IND-AD with empty BD vector and ETT-BD with empty AD vector) are shown. (E–H) FRET/FLIM analysis in Arabidopsis protoplast of IND-YFP (E) and ETT-CFP (F) and the merged signal in the nucleus (G–H). (I) Quantification of CFP lifetime with ETT-CFP and empty YFP vector (left) and ETT-CFP with IND-YFP (right). Bars: A–C, 50 µm.

Figure 4.

ETT and IND protein interact in an IAA-sensitive manner. (A–D) BiFC with pYFPN-ETT and pYFPC-IND in the presence of IAA (B), NAA (C), and 2,4-D (D). All hormonal treatments were with 1 mM in lanolin. (E) Fluorescence quantification (CTCF) of split YFP signal between ETT and IND without treatment and with IAA, NAA, and 2,4-D. (*) P < 0.01; (***) P < 0.0001. Error bars show the standard deviation. (F) Y2H assays with increasing concentration of IAA, NAA, and 2,4D. Bars: A–D, 50 µm.

Figure 5.

Auxin sensitivity is required for proper gynoecium morphogenesis. (A) Schematic representation of protein domains of IND and ETT. Numbers indicate amino acid positions. (B) Y2H assays of IND/IND^D30G versions with ETT-BD and ETT/ETT^2C-S versions with IND-AD and their sensitivity to IAA. Controls of IND^D30G, ETT^2C-S, and empty plasmids are shown below. (C–E) Phenotypic analyses of ind-2 (C), pIND::IND^WT in ind-2 (D), and pIND::IND^D30G in ind-2 (E) with SEM images of stage 13 gynoecia (panel I), dehiscence (panel II), pollen tube growth (panel III), and a seed set (panel IV). (F–H) SEM images of stage 13 gynoecia from ett-3 (F), pETT::ETT^WT in ett-3 (G), and pETT::ETT^2C-S in ett-3 (H). (stg) Stigma; (sty) style; (va) valve. Bars, 100 μm.

Figure 6.

Auxin sensitivity is required for proper organ morphogenesis. (A) Y2H assay showing IAA sensitivity of interactions of ETT-BD and TFs (fused to the AD domain) identified in the REGIA (Regulatory Gene Initiative in Arabidopsis) library. Assays were carried out in the absence of IAA (−IAA; top) and presence of IAA (+IAA; bottom). (B–D) Chart summarizing the comparison of LR numbers among the genotypes (normalized against root length), showing a statistically higher number of LR in the pETT::ETT^2C-S line (shown in B). GUS staining of the pETT(8Kb)::GUS line in the root (C) and LR defects in pETT::ETT^2C-S (D), with two adjacent LRs developing (indicated by asterisks). (E) pETT::ETT-GFP expression profile in ovule primordia at stage 2-II; the signal is localized predominantly in the region from which the integuments will develop. (F–H) Clearing of wild-type (F), ett-3 (G), and pETT::ETT^2C-S (H) ovules showing abnormal overgrowth of the outer integument in pETT::ETT^2C-S. (I,J) Stage 17 fruits from wild type (I) and pETT::ETT^2C-S (J), with one valve removed to expose the defect in fertility in the pETT::ETT^2C-S line. (K–N) GUS staining of the pETT(8Kb)::GUS line during lateral branch formation (K) and close-up views of the primary branch–stem internode in wild type (L), ett-3 (M), and pETT::ETT^2C-S (N). (O) Schematic model illustrating how the IAA-induced change in the ETT/TF complex dimerization state determines the regulation of downstream targets. As shown for PID in this study, transcriptomes A and B share genes regulated differently by the repressor/activator states. (nu) Nucellus; (ii) inner integument; (oi) outer integument; (fg) female gametophyte; (f) funiculus. Bars: C–H, 20 µm; L–M, 5 mm.