The E3 ligase ABI3-INTERACTING PROTEIN2 negatively regulates FUSCA3 and plays a role in cotyledon development in Arabidopsis thaliana

Abstract

FUSCA3 (FUS3) is a short-lived B3-domain transcription factor that regulates seed development and phase transitions in Arabidopsis thaliana. The mechanisms controlling FUS3 levels are currently poorly understood. Here we show that FUS3 interacts with the RING E3 ligase ABI3-INTERACTING PROTEIN2 (AIP2). AIP2-green fluorescent protein (GFP) is preferentially expressed in the protoderm during early embryogenesis, similarly to FUS3, suggesting that their interaction is biologically relevant. FUS3 degradation is delayed in the aip2-1 mutant and FUS3-GFP fluorescence is increased in aip2-1, but only during mid-embryogenesis, suggesting that FUS3 is negatively regulated by AIP2 at a specific time during embryogenesis. aip2-1 shows delayed flowering and therefore also functions post-embryonically to regulate developmental phase transitions. Plants overexpressing FUS3 post-embryonically in the L1 layer (ML1p:FUS3) show late flowering and other developmental phenotypes that can be rescued by ML1p:AIP2, further supporting a negative role for AIP2 in FUS3 accumulation. However, additional factors regulate FUS3 levels during embryogenesis, as ML1:AIP2 seeds do not resemble fus3-3. Lastly, targeted expression of a RING-inactive AIP2 variant to the protoderm/L1 layer causes FUS3 and ABI3 overexpression phenotypes and defects in cotyledon development. Taken together, these results indicate that AIP2 targets FUS3 for degradation and plays a role in cotyledon development and flowering time in Arabidopsis.

Keywords: AIP2; E3 ligase; FUSCA3; embryogenesis; post-translational regulation; protein degradation; protein localization; protein–protein interaction; seed development; transcription factor..

Fig. 1.

AIP2 interacts with FUS3 in yeast-two hybrid assay, in vitro, and in planta. (A) Schematic diagram of FUS3 and AIP2 variants used in Y2H and pull-down assays. The B2 and B3 basic domains, the activation (A) domain, and the PEST degradation motif of FUS3 are shown. The location of C230S, C231S, and E235G mutations in AIP2^{(C/S, E/G)} are indicated by the triangle. Numbers refer to the amino acid residues in the proteins. (B) Y2H assays showing interaction between BD-FUS3^(N90) and AD-AIP2 variants on selective media. The empty pJG4-5 vector containing the B42 activation domain (AD) was used as the negative control. BD, LexA DNA-binding domain. (C) In vitro pull-down assays of GST–FUS3 (~66 kDa) with AIP2-6×His and inactive AIP2^(C/S,E/G)-6×His (~40 kDa). Immunoblots using anti-His and anti-GST antibodies show interaction of AIP2 and AIP2^{(C/S, E/G)} with FUS3. GST was used as the negative control. (D) In vitro pull-down assay showing interaction of FUS3 with AIP2 and its variants. GST–FUS3 (2.5 µg) was incubated with 35Sp:HA-AIP2 or 35Sp:HA-AIP2^C/S,E/G plant cell lysates, pulled-down using glutathione resin, and detected with anti-GST antibody. The AIP2 and AIP2^C/S,E/G were detected with anti-HA antibody. I, input protein sample. (E) Confocal images showing interaction between FUS3 and AIP2 by BiFC in N. benthamiana. Transient co-expression of nYFP–FUS3 with cYFP–AIP2 showing YFP fluorescence in the nucleus and cytoplasm of pavement cells. No fluorescence was detected when the negative control, nYFP–MYB49, was co-expressed with cYFP–AIP2. The same confocal settings were used in both images. Scale bars=50 ìm. (This figure is available in colour at JXB online.)

Fig. 2.

Expression pattern of AIP2 during embryogenesis and in seedlings. (A) FUS3 and AIP2 mRNA levels during embryogenesis (BAR; Toufighi et al., 2005). (B) Confocal images of AIP2p:GFP embryos (top panels). Bottom panels show paradermal optical sections of walking-stick and bent-cotyledon embryos. (C) Confocal images of AIP2p:GFP seedlings. Top panels, paradermal optical sections of cotyledons; bottom panels, paradermal (left) and median longitudinal (right) optical sections of leaves of seedlings 7 days after imbibition (DAI). Scale bars=50 ìm. All images were taken under comparable confocal settings. (This figure is available in colour at JXB online.)

Fig. 3.

FUS3 and AIP2 protein localization patterns during embryogenesis and germination. (A) Confocal images showing GFP fluorescence in the protoderm of fus3-3 FUS3p:FUS3-GFP and aip2-1 AIP2p:GFP-AIP2 embryos at walking-stick and bent-cotyledon stages of embryogenesis. (B) Confocal images of cotyledons, hypocotyls, and roots of FUS3p:FUS3-GFP and AIP2p:GFP-AIP2 seedlings during germination. Longitudinal optical sections of the root and paradermal optical sections of the cotyledon and hypocotyl are shown. All images were taken under comparable confocal settings. Images shown were merged combining the GFP fluorescence and autofluorescence from chlorophyll. Scale bars=20 μm. (C) Median-longitudinal section of hypocotyls of an embryo 10 d after fertilization showing preferential localization of GFP–AIP2 to the protoderm. (This figure is available in colour at JXB online.)

Fig. 4.

Lack of AIP2 increases the FUS3–GFP level during mid-embryogenesis. Confocal images (A) and quantification (B) showing increased FUS3–GFP fluorescence in aip2-1 fus3-3 FUS3p:FUS3-GFP compared with fus3-3 FUS3p:FUS3-GFP walking-stick embryos. All images were taken under comparable confocal settings. Images shown were merged combining the GFP fluorescence and autofluorescence from chlorophyll. Scale bar=20μm. (*) p < 0.05 (two-tailed t-test). (This figure is available in colour at JXB online.)

Fig. 5.

AIP2 negatively regulates FUS3. (A) Immunoblots of cell-free degradation assays showing glutathione S-transferase (GST)–FUS3 protein levels after incubation in wild-type (WT) and aip2-1 lysates for 60 min. FUS3–GST was detected with anti-GST antibody. Protein lysates were extracted from seedlings at 7 days after imbibition (DAI). Three experiments were conducted and one representative is shown. (B) Degradation kinetics of GST–FUS3. Averages of three experiments ±SD are shown.

Fig. 6.

Rescue of ML1p:FUS3 phenotypes by ML1p:AIP2. (A) ML1p:FUS3-GFP late flowering (top panel), aborted siliques (arrow heads; middle panel), and glabrous leaves (arrows; bottom panel) are rescued in ML1p:FUS3-GFP ML1p:HA-AIP2 double transgenic plants. Plants were grown for 1 month on soil under a long-day growth cycle. (B) Quantification of flowering time (days to flowering) for genotypes shown in (A). n=10 plants for all genotypes. Only 5 out of 10 ML1p:FUS3-GFP plants flowered; *P<0.0002 from the wild type (WT); **P<0.02 from ML1p:FUS3-GFP. (C) Immunoblot of seedlings at 4 d after imbibition of two independent ML1p:HA-AIP2 ML1p:FUS3-GFP double transgenic plants. The blot was probed with GFP antibody to detect FUS3–GFP (~60 kDa). (*) cross-reacting bands. (This figure is available in colour at JXB online.)

Fig. 7.

Phenotype of ML1p:HA-AIP2^(C/S,E/G) and model of the role of AIP2 in plant development. (A) Phenotypes of seedlings from two ML1p:HA-AIP2^(C/S,E/G) lines, showing defects in cotyledon development, including reduced or lack of chlorophyll, altered cotyledon number (monocotyledons and polycotyledons), and altered morphology (narrow cotyledons, cotyledon fusion). Arrows point to glabrous leaves or leaves showing few, unbranched trichomes, resembling ML1p:FUS3 leaves, compared with the wild type (WT) (a close-up image of the WT is shown in the bottom, left corner panel). (B) Model showing the role of AIP2 as a negative regulator of FUS3 and ABI3. Other unknown negative regulators (such as E3s) or other mechanisms also control FUS3 and ABI3 protein levels. AIP2 may regulate cotyledon development through auxin. (This figure is available in colour at JXB online.)