In Silico Characterisation of the Late Embryogenesis Abundant (LEA) Protein Families and Their Role in Desiccation Tolerance in Ramonda serbica Panc

Abstract

Ramonda serbica Panc. is an ancient resurrection plant able to survive a long desiccation period and recover metabolic functions upon watering. The accumulation of protective late embryogenesis abundant proteins (LEAPs) is a desiccation tolerance hallmark. To propose their role in R. serbica desiccation tolerance, we structurally characterised LEAPs and evaluated LEA gene expression levels in hydrated and desiccated leaves. By integrating de novo transcriptomics and homologues LEAP domains, 318 R. serbica LEAPs were identified and classified according to their conserved motifs and phylogeny. The in silico analysis revealed that hydrophilic LEA4 proteins exhibited an exceptionally high tendency to form amphipathic α-helices. The most abundant, atypical LEA2 group contained more hydrophobic proteins predicted to fold into the defined globular domains. Within the desiccation-upregulated LEA genes, the majority encoded highly disordered DEH1, LEA1, LEA4.2, and LEA4.3 proteins, while the greatest portion of downregulated genes encoded LEA2.3 and LEA2.5 proteins. While dehydrins might chelate metals and bind DNA under water deficit, other intrinsically disordered LEAPs might participate in forming intracellular proteinaceous condensates or adopt amphipathic α-helical conformation, enabling them to stabilise desiccation-sensitive proteins and membranes. This comprehensive LEAPs structural characterisation is essential to understanding their function and regulation during desiccation aiming at crop drought tolerance improvement.

Keywords: 3D protein structure modelling; de novo transcriptome assembly; differentially expressed gene analysis; drought; intrinsically disordered proteins; liquid–liquid phase separation; resurrection plants; secondary structure prediction.

Figure 1

Percentage of selected amino acids and Gly versus GRAVY index plot in R. serbica LEA protein family members. The distribution of hydrophilins is highlighted in grey in the Gly/GRAVY plot.

Figure 2

MEME motifs and motif logos of representative LEAPs of each R. serbica LEA protein family group and subgroup. The numbers in the parentheses present the RsLEAP code (see Supplementary Tables S2 and S3). The consensus and logo sequences of each motif are presented in Table 3. The numbers at the end of each protein sequence present a percentage of LEAPs with the same motif pattern in the respective LEA protein family group. The bar represents 100 aa.

Figure 3

Average distribution of the predicted secondary structure of each R. serbica LEA protein family group and subgroup members according to the five secondary structure prediction algorithms: PsiPred, Sopma, FELLS, Phyre2, and JPred4.

Figure 4

Modelling of the α-helix structure within detected MEME motifs in R. serbica LEA protein family members. Helical projections of α-helices were generated using the HeliQuest webserver [47]. αH; predicted α-helix percentages obtained by FELLS. Each wheel was obtained with an 11-amino acid window. The arrow shows the helical hydrophobic moment.

Figure 5

Three-dimensional models of the representative LEAPs of each R. serbica LEA protein family group. Detected MEME motifs are denoted in blue. The RsLEA code for each protein is given. Orange: transmembrane α-helix, TMH.

Figure 6

Repartition of the R. serbica LEAP-predicted subcellular distribution in each family (sub)group. Results are given in percentages. Chl, chloroplast; mito, mitochondrion; nucl, nucleus; pero, peroxisomes; cysk, cytoskeleton; cyto, cytosol; ext, extracellular space; golg, Golgi apparatus, E.R. endoplasmic reticulum; plas, plastids.