Inflorescence deficient in abscission controls floral organ abscission in Arabidopsis and identifies a novel family of putative ligands in plants

Abstract

Abscission is an active process that enables plants to shed unwanted organs. Because the purpose of the flower is to facilitate pollination, it often is abscised after fertilization. We have identified an Arabidopsis ethylene-sensitive mutant, inflorescence deficient in abscission (ida), in which floral organs remain attached to the plant body after the shedding of mature seeds, even though a floral abscission zone develops. The IDA gene, positioned in the genomic DNA flanking the single T-DNA present in the ida line, was identified by complementation. The gene encodes a small protein with an N-terminal signal peptide, suggesting that the IDA protein is the ligand of an unknown receptor involved in the developmental control of floral abscission. We have identified Arabidopsis genes, and cDNAs from a variety of plant species, that encode similar proteins, which are distinct from known ligands. IDA and the IDA-like proteins may represent a new class of ligands in plants.

Figure 1.

Phenotype of the ida Mutant Compared with the Wild Type and etr1-1. (A) Inflorescences of wild-type (wt) Arabidopsis and the ida mutant. The arrowhead indicates the first wild-type flower position where the floral organs have abscised. Note ida flowers attached along the entire inflorescence. (B) Flowers along the inflorescence. Note the indications of senescence in ida flowers from position 16, showing that the mutant is sensitive to ethylene. Note also fresh, turgid, and vivid floral organs at all positions before abscission in the ethylene-insensitive etr1-1 mutant. (C) Septum from a dry plant after dehiscence of seeds with dry floral organs still attached in the ida mutant, in contrast to the wild type.

Figure 2.

Morphology of AZs and Petal Breakstrength. Scanning electron microscopy of the fracture plane AZ cells of petals in wild-type ([A] to [E]) and ida ([F] to [J]) flowers at positions 4 ([A] and [F]), 8 ([B] and [G]), 10 ([C] and [H]), 12 ([D] and [I]), and 22 ([E] and [J]) after forcible removal or natural abscission. Wild-type petals abscise naturally at position 8. (K) shows petal breakstrength (i.e., the force required to remove the petals from the flowers) measured from positions 2 to 32 along the inflorescence. Breakstrength was measured for 15 wild-type and ida plants, with a minimum of 20 measurements at each position. Standard deviations are shown as thin lines at the top of the columns.

Figure 3.

Ethylene Response of the ida Mutant. (A) The triple-response assay was conducted by exposing seedlings growing on vertical half-strength Murashige and Skoog (1962) agar plates to 10 ppm ethylene for 3 days in the dark. Note that ida reacts as a wild-type (wt) seedling, whereas etr1-1 is insensitive to ethylene and behaves like a control seedling subjected to air. (B) Mature wild-type, ida, and etr1-1 adult plants were exposed to 10 ppm ethylene for 48 h in flow-through chambers. Note the yellow, wilting rosette leaves of the wild type and ida but not of etr1-1. (C) Comparison of flowers at positions 0 to 6 exposed to air or to 10 ppm ethylene. In wild-type flowers, exposure to ethylene causes senescence and floral organ abscission already in flowers at position 1. By contrast, ida flowers senesce but do not abscise.

Figure 4.

Identification of the IDA Gene. (A) The single T-DNA with 1239 bp of vector backbone was inserted in chromosome 1 between genes At1g68780 and At1g68765 in the ida line. The T-DNA insertion had resulted in a 74-bp target-site deletion (Δ). The distances from the stop codon and the start codon of these two genes, respectively, to the T-DNA insertion point are shown (bp), as are the numbers of base pairs from the start to the stop codons in the two genes. The single intron in At1g68780 is indicated in black. The direction of transcription is indicated by arrows. The extents of the two genomic fragments used in a complementation experiment are shown below the genomic region. All elements in the drawing are not to scale. (B) The phenotype of ida plants transformed with the fragment encompassing the At1g68780 gene was identical to the ida mutant phenotype (a). Note flowers attached to full-grown siliques (b). (C) The phenotype of ida plants transformed with the fragment covering the At1g68780 gene was identical to the wild-type phenotype (a). Note that floral organs are absent from siliques in the lower positions of the inflorescence (b). (D) RT-PCR on mRNA from flowers at positions 1 to 8 amplified the expected 421-bp fragment with IDA primers from wild-type (wt) but not from ida plants, whereas a 294-bp ACTIN2-7 fragment was amplified from both wild-type and ida plants. M, φX174 digested with HaeIII.

Figure 5.

IDA:GUS Expression Assay. Cleared whole-mount preparations of GUS-stained IDA:GUS flowers from different developmental stages. (A) Developmental assay showing stage-specific AZ GUS expression in early IDA:GUS flower stages. Top row, whole-flower overview; bottom row, AZ detail. Arabic numerals indicate flower positions on the inflorescence. (B) and (C) AZ GUS expression in flowers at positions 10 and 15. The arrowhead indicates nectary outgrowth. S, mature seed. (D) GUS expression in the AZ (stippled lines) and in nectaries (arrowheads) between the sepal bases (se) and anther filament bases (af). O, ovule. (E) GUS expression in the AZ and in the anther filament base. (F) Detail of (E). Note the GUS expression in the anther filament AZs on both the plant body side and the anther filament side. The arrowhead marks nectary outgrowth with weak GUS expression. (G) IDA:GUS expression in the sepal. The signal seems to diffuse from the base toward the sepal apex.

Figure 6.

Subcellular Localization of Proteins in an Onion Epidermis Transient Expression Assay. GFP fluorescence was revealed 2 h after bombardment using a Nikon Microphot microscope (Tokyo, Japan) equipped with epifluorescence and Nomarski optics. (A) Fluorescence micrographs of the extracellular localization of the IDA:GFP fusion protein. Note the presence of GFP signal around several neighboring cells. (B) Fluorescence micrographs of the extracellular localization of the IDA signal peptide:GFP fusion protein. Note the presence of GFP signal around several neighboring cells. (C) Fluorescence micrographs of the cytoplasmic localization of GFP alone. The signal in the nucleus is the result of diffusion of the small GFP protein (∼25 kD) (von Arnim et al., 1998). Note that the GFP signal is not seen in neighboring cells. (D) Merged fluorescence and differential interference contrast micrographs of the nuclear localization of the GFP:HP1 fusion protein, a control for subcellular localization. Note that the GFP signal is not seen in neighboring cells.

Figure 7.

Proteins Encoded by IDL Genes and Expression Patterns in Arabidopsis. (A) Alignment of IDA and IDL proteins encoded by cDNAs from Arabidopsis (AtIDL1), tomato (LeIDL1), lotus (LjIDL1), soybean (GmIDL1), black locust (RpIDL1), maize (ZmIDL1), poplar (PtIDL1), wheat (TaIDL1), and four putative IDL genes of Arabidopsis (see Table 1). Note the hydrophobic predicted signal peptides and the arrow indicating the positions of cleavage sites predicted by SignalP. Amino acids are shaded according to properties: light gray with dark letters, hydrophobic residues; dark gray with dark letters, basic residues; light gray with white letters, small residues; dark gray with white letters, polar nonaliphatic residues. The best conserved residues are shown below the alignment: uppercase letters, amino acids identical in all proteins; lowercase letters, amino acids identical in at least eight proteins; 6, hydrophobic residue (Val or Ile); 4, positive residue (Lys or Arg). The conserved PIP motif is indicated. (B) RT-PCR on mRNA from different tissues, as indicated (bottom gels), using primers amplifying fragments of 465 bp (AtIDL1), 280 bp (AtIDL2), 256 bp (AtIDL3), 261 bp (AtIDL4), and 259 bp (AtIDL5). At top, genomic DNA was used as a template, and the marker line (M) is shown. Note that the positive control for ACTIN2-7, giving a fragment of 255 bp with primers spanning intron 2, was amplified at comparable levels from all tissues. Note also that the negative control PCR gel—ACTIN (no RT)—shows no products except for the ACTIN2-7 genomic control of 340 bp.