Molecular and phylogenetic characterization of the sieve element occlusion gene family in Fabaceae and non-Fabaceae plants

Abstract

Background: The phloem of dicotyledonous plants contains specialized P-proteins (phloem proteins) that accumulate during sieve element differentiation and remain parietally associated with the cisternae of the endoplasmic reticulum in mature sieve elements. Wounding causes P-protein filaments to accumulate at the sieve plates and block the translocation of photosynthate. Specialized, spindle-shaped P-proteins known as forisomes that undergo reversible calcium-dependent conformational changes have evolved exclusively in the Fabaceae. Recently, the molecular characterization of three genes encoding forisome components in the model legume Medicago truncatula (MtSEO1, MtSEO2 and MtSEO3; SEO = sieve element occlusion) was reported, but little is known about the molecular characteristics of P-proteins in non-Fabaceae.

Results: We performed a comprehensive genome-wide comparative analysis by screening the M. truncatula, Glycine max, Arabidopsis thaliana, Vitis vinifera and Solanum phureja genomes, and a Malus domestica EST library for homologs of MtSEO1, MtSEO2 and MtSEO3 and identified numerous novel SEO genes in Fabaceae and even non-Fabaceae plants, which do not possess forisomes. Even in Fabaceae some SEO genes appear to not encode forisome components. All SEO genes have a similar exon-intron structure and are expressed predominantly in the phloem. Phylogenetic analysis revealed the presence of several subgroups with Fabaceae-specific subgroups containing all of the known as well as newly identified forisome component proteins. We constructed Hidden Markov Models that identified three conserved protein domains, which characterize SEO proteins when present in combination. In addition, one common and three subgroup specific protein motifs were found in the amino acid sequences of SEO proteins. SEO genes are organized in genomic clusters and the conserved synteny allowed us to identify several M. truncatula vs G. max orthologs as well as paralogs within the G. max genome.

Conclusions: The unexpected occurrence of forisome-like genes in non-Fabaceae plants may indicate that these proteins encode species-specific P-proteins, which is backed up by the phloem-specific expression profiles. The conservation of gene structure, the presence of specific motifs and domains and the genomic synteny argue for a common phylogenetic origin of forisomes and other P-proteins.

Figure 1

Maximum likelihood phylogenetic tree of SEO proteins from different plants. The phylogenetic tree was constructed with RAxML from a T-Coffee protein sequence alignment and with a bootstrap support of 1000 replicates. Bootstrap percentages are shown on the nodes. Branch lengths are proportional to the number of amino acid substitutions. The shaded parts of the tree represent the subgroups identified with OrthoMCL. Mt = Medicago truncatula, Gm = Glycine max, Vf = Vicia faba, Cg = Canavalia gladiata, Ps = Pisum sativum, Md = Malus domestica, At = Arabidopsis thaliana, Vv = Vitis vinifera, Sp = Solanum phureja, pot. ψ = potential pseudogene, pot. ψe = expressed potential pseudogene.

Figure 2

Schematic overview of SEO gene exon-intron structure. Exons are represented as colored boxes, introns as dashed lines. Introns are not drawn to scale, but the length of the individual introns is indicated in base pairs. Mt = Medicago truncatula, Gm = Glycine max, Md = Malus domestica, At = Arabidopsis thaliana, Sp = Solanum phureja.

Figure 3

RT-PCR analysis of SEO gene expression in M. truncatula, G. max, A. thaliana and S. phureja. SEO genes were amplified from cDNA prepared from total RNA isolated from phloem-enriched (PE) and phloem-deficient (PD) tissue. The constitutively expressed ACT2 (for A. thaliana), GAPDH (for M. truncatula, S. phureja) and F-box (for G. max) genes served as positive controls. The integrity of all PCR products was verified by sequencing.

Figure 4

Analysis of the AtSEOa and GmSEO-F1 promoter in transgenic plant tissue. CLSM detection of GFP_ER activity in PAtSEOa-GFP_ER transgenic A. thaliana plants (A to D) and PGmSEO-F1-GFP_ER transgenic G. max roots (E, F). (A) Transverse sections through an A. thaliana stem showing GFP_ER restricted to the phloem. (B) Overlay of fluorescent and transmitted light images of a longitudinal A. thaliana stem section showing GFP_ER fluorescence in sieve elements (arrows). Non-fluorescent companion cells are marked with an asterisk. (C) Sieve plate of two end-to-end connected fluorescent sieve elements stained with aniline-blue. (D) Sieve element containing large vacuoles, indicated by the white arrow. (E) Longitudinal section through the vascular cylinder of a transgenic G. max root. (F) Sieve element with aniline-blue stained sieve plate. Scale bar = 100 μm in A and E, and 5 μm in B-D and F.

Figure 5

Predicted thioredoxin fold of MtSEO-F1 and AtSEOa. Three-dimensional structure of the potential thioredoxin fold in (A) MtSEO-F1 and (B) AtSEOa predicted with I-TASSER. α-helices are coloured in red, β-sheets in blue and turns in yellow. Comparison of the predicted three-dimensional structures of (C) MtSEO-F1 and (D) AtSEOa (both coloured in red) with tryparedoxin II, aligned with TM-align. The structure of tryparedoxin II is coloured in blue, cysteine residues are highlighted in green.

Figure 6

Domain arrangement of the expressed SEO proteins. (A) The phylogenetic tree was calculated with RaxML from a T-Coffee protein alignment. Identified protein domains are drawn to scale. The positions of the SEO-NTD and SEO-CTD domains and the potential thioredoxin fold are indicated. (B) Sequence logos: M1 = sequence logo specific for all SEO proteins; M2 = sequence logo specific for SEO subgroups 1-3; M3 = sequence logo specific for SEO subgroups 5, 6 and 7; and M4 = sequence logo specific for SEO subgroups 5 and 6, with potential intrinsic disorder. The position of the individual sequence logos is also indicated in (A).

Figure 7

Genomic synteny of several SEO genes from M. truncatula and G. max. Schematic overview of the genomic synteny of SEO clusters on M. truncatula chromosome 1 and G. max chromosomes 2, 10, 13 and 20. SEO genes are shown as red arrows. Other gene models are shown as black arrows. Orthologs between the two Fabaceae and paralogs within G. max are connected by dark grey shading. Orthologs and paralogs for non-SEO genes are indicated by light grey shading.

Figure 8

Phylogenetic tree of the plants included in this investigation. Phylogenetic tree of the plants included in this investigation according to the Angiosperm Phylogeny Group [75]. Clades and families are shown within the tree, and the numbers of known SEO genes are indicated.