Self-association and multimer formation in AtLEA4-5, a desiccation-induced intrinsically disordered protein from plants

Abstract

During seed maturation, plants may experience severe desiccation, leading to the accumulation of late embryogenesis abundant (LEA) proteins. These intrinsically disordered proteins also accumulate in plant tissues under water deficit. Functional roles of LEA proteins have been proposed based on in vitro studies, where monomers are considered as the functional units. However, the potential formation of homo-oligomers has been little explored. In this work, we investigated the potential self-association of Arabidopsis thaliana group 4 LEA proteins (AtLEA4) using in vitro and in vivo approaches. LEA4 proteins represent a compelling case of study due to their high conservation throughout the plant kingdom. This protein family is characterized by a conserved N-terminal region, with a high alpha-helix propensity and invitro protective activity, as compared to the highly disordered and low-conserved C-terminal region. Our findings revealed that full-length AtLEA4 proteins oligomerize and that both terminal regions are sufficient for self-association in vitro. However, the ability of both amino and carboxy regions of AtLEA4-5 to self-associate invivo is significantly lower than that of the entire protein. Using high-resolution and quantitative fluorescence microscopy, we were able to disclose the unreported ability of LEA proteins to form high-order oligomers in planta. Additionally, we found that high-order complexes require the simultaneous engagement of both terminal regions, indicating that the entire protein is needed to attain such structural organization. This research provides valuable insights into the self-association of LEA proteins in plants and emphasizes the role of protein oligomer formation.

Keywords: Arabidopsis thaliana; desiccation; intrinsically disordered proteins; late embryogenesis abundant proteins; oligomerization; protein–protein interaction; water deficit.

FIGURE 1

Self‐association of LEA4 recombinant proteins. (a, b) Top panels represent AtLEA4‐2 and AtLEA4‐5 proteins, respectively. Color rectangles indicate the terminal regions, showing their length by the amino acid location from start to end. Blue and red rectangles represent the N‐ and the C‐terminal regions, respectively. Middle panels display the protein structural disorder predicted using Metapredict V2. Bottom panels represent the distribution of helicity throughout the proteins, predicted with FELLS. (c, d) SDS‐PAGE of recombinant AtLEA4‐2 (10.5 kDa), and AtLEA4‐5 proteins (16.2 kDa), respectively, indicating their apparent molecular mass. Because its small size, AtLEA4‐2 was separated in SDS‐PAGE using Tris‐tricine buffer, whereas for AtLEA4‐5, we used the Tris‐glycine system. Arrowheads indicate the bands analyzed by mass spectrometry for identification. Molecular mass markers (kDa) are displayed on the left. (e) Separation of AtLEA4‐5 protein (50 μg) in a native 10% polyacrylamide gel electrophoresis showing oligomeric forms. Three globular proteins with similar isoelectric points to AtLEA4‐5 (pI = 9.5) were used as reference of migration: Cytochrome C (Cyt C, 13.4 KDa, pI = 9.2), Lysozyme (Lys, 14.4 kDa, pI = 9.03), and Ribonuclease (RNAse, 13.7 kDa, pI = 8.22). (f) Elution profile of purified AtLEA4‐5 protein (2.35 mM, 0.1 mL). AtLEA4‐5 monomer and dimer are indicated as M and D, respectively. The elution of the molecular mass markers is indicated by red dotted lines, corresponding from left to right to bovine serum albumin (66.4 kDa), egg albumin (45 kDa), and carbonic anhydrase (28 kDa). (g) Analysis of SEC fractions corresponding to the major peaks were separated in SDS‐PAGE (15% polyacrylamide) and analyzed by Western blot. The retention volumes of the fractions selected for gel loading are indicated in the upper part of the gel. AtLEA4‐5 monomer and dimer are indicated as M and D, respectively. the 8 mL fraction just showed a band corresponding to the AtLEA4‐5 dimer.

FIGURE 2

Multimers and high‐order particles of AtLEA4‐5 detected by dynamic light scattering. (a) Size distribution by intensity percentage of different concentrations of AtLEA4‐5 in 25 mM HEPES buffer pH 7.5. The basal intensity was the same for all curves. The intensity of each curve was increased by two units to facilitate the comparison. The filling transparency for each protein concentration represents the standard deviation of five technical repetitions. (b) Hydrodynamic radius (Rh) obtained from different concentrations of AtLEA4‐5 solutions. Rh was estimated by the Stokes–Einstein equation, using the diffusion coefficient obtained from the autocorrelation curve. The filling transparency represents the standard deviation of five technical repeats.

FIGURE 3

In vitro cross‐linking detects LEA4 discrete oligomeric protein assemblies. (a, b) Western blot analysis of photo‐cross‐linked AtLEA4‐2 and AtLEA4‐5 recombinant proteins, respectively. Proteins were separated in 10% SDS‐PAGE in a Tris‐glycine system. The crosslinking assay was performed with increasing protein concentrations, as indicated. (c, d) Western blots of cross‐linked AtLEA4‐5_1‐77 and AtLEA4‐5_78‐158 recombinant proteins, respectively, using specific antibodies for each (c) were separated in a gradient polyacrylamide gel 6–18%, whereas for proteins in (d), 10% SDS‐PAGE in a Tris‐tricine system was used to enhance the resolution in the separation of the protein complexes.

FIGURE 4

In‐depth tissue‐level oligomerization analysis of AtLEA4‐5 using BiFC assays. (a) Schematic representation of the constructs used for BiFC analysis. From top to bottom: Full‐length AtLEA4‐5 constructs: pYFN‐4‐5 and pYFC4‐5, AtLEA4‐5 amino‐end constructs: pYFN‐4‐5_1‐77 and pYFC‐4‐5_1‐77, AtLEA4‐5 carboxy‐end constructs: pYFN‐4‐5_78‐158 and pYFC‐4‐5_78‐158, and empty vector: pYFN and pYFC. (b) BiFC assay images of N. benthamiana leaves transiently transformed with the full AtLEA4‐5 protein (pYFN‐4‐5/pYFC‐4‐5), its N‐terminal (pYFN‐4‐5_1‐77/pYFC‐4‐5_1‐77), and C‐terminal (pYFN‐4‐5_78‐158/pYFC‐4‐5_78‐158) regions. As negative control, we used constructs with only the YFP fragment coding regions (pYFN and pYFC). All images are maxed intensity projections with a consistent brightness and contrast scale (range 250–9000 digital levels), visualized using a yellow lookup table (LUT). (c) Quantitative fluorescence intensity analysis, normalized to the empty vector (pYFN and pYFC) (see section 4 for details). Bars represent the relative fluorescence intensity, with error bars showing standard deviation. Each bar represents the relative fluorescence of a minimum of three plants for each construct, assessing up to three leaves per plant and evaluating over 50 volumetric scenes of 100 × 100 × 25 mm³ each (details in Figure S4). Individual data points overlaid on the bar graph depict the mean relative fluorescence of experiments shown in Figure S4. Statistically significant differences between constructs were determined using an ANOVA test, yielding a p‐value of approximately 0.002. Post hoc pairwise comparisons between groups were conducted using the Games‐Howell test to identify significant differences between constructs. Significance levels are denoted as follows: *p‐value <0.05, ** p‐value <0.01, and “n.s.” nonsignificant differences.

FIGURE 5

In vivo single‐cell and subcellular oligomerization analysis of AtLEA4‐5 protein. (a) Fluorescence intensity analysis of AtLEA4‐5 variants at the single‐cell level, showing correspondence with the tissue‐level observations. BiFC assays in N. benthamiana leaves were visualized using the Green Fire Blue lookup table in Image J (version 1.55f). Displayed variants encompass the full AtLEA4‐5 protein (pYFN‐4‐5 and pYFC‐4‐5), its N‐terminal (pYFN‐4‐5_1‐77 and pYFC‐4‐5_1‐77), C‐terminal (pYFN‐4‐5_78‐158 and pYFC‐4‐5_78‐158), and negative controls with only the YFP fragment coding regions (pYFN and pYFC). Insets offer a normalized, autoscaled perspective for clarity. (b) Brightness evaluations of the full and truncated AtLEA4‐5 versions at the subcellular level, emphasizing the distinct variations among constructs visualized using the Royal lookup table in Image J (version 1.55f). (c) Brightness comparison of the entire AtLEA4‐5 protein, its amino‐terminal, and carboxy‐terminal regions in segmented individual cells, using a threshold of B = 5. This threshold delineates the transition from dimers (depicted in green) to oligomers (shown in red) based on stoichiometry. The lower panels represent the distribution of monomers to dimers, expressed as percentages, from a sample of up to 42 cells for each AtLEA4‐5 variant. Detailed single‐cell examples are provided in Figure S5. (d) Correlation plot showcasing the relationship between fluorescence intensity and brightness for the full AtLEA4‐5 protein and its truncated counterparts. Each dot represents the mean intensity and brightness measured at the single‐cell level. (e) Diffusion coefficients for the AtLEA4‐5 protein and its variants, elucidating the interplay between oligomerization and subcellular molecular mobility. Willcoxon Rank sum exact test was applied for this analysis, *p‐value <0.05, **p‐value <0.01. Collectively, the data underscore the AtLEA4‐5 protein's propensity to form high‐order oligomers within individual plant cells.