Partial functional conservation of IRX10 homologs in physcomitrella patens and Arabidopsis thaliana indicates an evolutionary step contributing to vascular formation in land plants

Abstract

Background: Plant cell walls are complex multicomponent structures that have evolved to fulfil an essential function in providing strength and protection to cells. Hemicelluloses constitute a key component of the cell wall and recently a number of the genes thought to encode the enzymes required for its synthesis have been identified in Arabidopsis. The acquisition of hemicellulose synthesis capability is hypothesised to have been an important step in the evolution of higher plants.

Results: Analysis of the Physcomitrella patens genome has revealed the presence of homologs for all of the Arabidopsis glycosyltransferases including IRX9, IRX10 and IRX14 required for the synthesis of the glucuronoxylan backbone. The Physcomitrella IRX10 homolog is expressed in a variety of moss tissues which were newly formed or undergoing expansion. There is a high degree of sequence conservation between the Physcomitrella IRX10 and Arabidopsis IRX10 and IRX10-L. Despite this sequence similarity, the Physcomitrella IRX10 gene is only able to partially rescue the Arabidopsis irx10 irx10-L double mutant indicating that there has been a neo- or sub-functionalisation during the evolution of higher plants. Analysis of the monosaccharide composition of stems from the partially rescued Arabidopsis plants does not show any significant change in xylose content compared to the irx10 irx10-L double mutant. Likewise, knockout mutants of the Physcomitrella IRX10 gene do not result in any visible phenotype and there is no significant change in monosaccharide composition of the cell walls.

Conclusions: The fact that the Physcomitrella IRX10 (PpGT47A) protein can partially complement an Arabidopsis irx10 irx10-L double mutant suggests that it shares some function with the Arabidopsis proteins, but the lack of a phenotype in knockout lines shows that the function is not required for growth or development under normal conditions in Physcomitrella. In contrast, the Arabidopsis irx10 and irx10 irx10-L mutants have strong phenotypes indicating an important function in growth and development. We conclude that the evolution of vascular plants has been associated with a significant change or adaptation in the function of the IRX10 gene family.

Figure 1

Sequence alignment of the Arabidopsis, Populus and Physcomitrella IRX10 protein family. IRX10 related protein sequences from Arabidopsis, Populus and Physcomitrella were aligned using Clustal-X. Shading indicates 100% conservation (black), 80% conservation (dark grey) or 60% conservation (light grey).

Figure 2

Phylogenetic trees for GX related proteins in Arabidopsis and Physcomitrella. Predicted protein sequences were aligned within each family, and a tree was computed using the neighbour-joining method [46] with gapped positions being excluded and branch length corrections for multiple substitutions enabled. The bars represent a PAM value (percent accepted mutations) of 10%. The numbers shown at the branch points are bootstrap values derived from 1000 randomized sequences. A) GT43 family. B) GT47 family. C) GT8 family.

Figure 3

The expression pattern of PpGT47A in Physcomitrella, analyzed by histochemical staining of stable pGT47A:GUS lines. (A) Schematic drawing describing the GUS construct inserted into the Physcomitrella genome. (B) Adult gametophore shows GUS staining in the apical region and in the emerging gametophore on the side of the adult gametophore. (C) GUS staining of the antheridia. (D) GUS staining in tissue below the immature sporophyte. (E) Expression peak in young gametophore growing on the side of old gametophore. (F) Protonema side branches undergoing tip growth exhibit peaks in GUS expression. (G) Bud/juvenile gametophore exhibiting GUS staining.

Figure 4

The Physcomitrella IRX10 homolog expressed in the Arabidopsis irx10 irx10-L double mutant. A) Arabidopsis irx10 irx10-L double mutant partially complemented by p35S:PpGT47A. B) Rosette diameter and inflorescence height of PpGT47A complemented Arabidopsis plants. Data obtained from between 2 and 7 biological repeats per line. Each line represents one independent transformation event. C) Sections from 8-week old basal stem tissue of Columbia wild-type, Arabidopsis irx10 irx10-L double mutant complemented with the Physcomitrella IRX10 homolog, and the Arabidopsis irx10 irx10-L double mutant. D) Higher magnification of the safranin stained tissues shown in C. Safranin staining in the cell corners of the irx10 irx10-L double mutant is indicated by arrowheads. Scale bar 50 μm (C) or 20 μm (D).

Figure 5

The PpGT47A gene does not rescue the monosaccharide composition of Arabidopsis irx10 irx10-L mutant stem tissue.A) Monosaccharide composition of stems material isolated from mature Arabidopsis wild-type plants, young wild-type plants with a 30mm stem, irx10 irx10-L double mutants and 3 independent lines of irx10 irx10-L double mutants complemented by the Physcomitrella IRX10 homolog (AtPpGT47A:16, AtPpGT47A:8, AtPpGT47A:5). Data from complemented plants are based on pooled stem material in three technical replicates for each of 3 independent biological samples pooled from multiple plants (except for samples indicated by * where only a single pooled sample was used). Data is displayed as molecular percentage of measured sugars obtained by TMS analysis from 500 μg of amylase treated AIR starting material for each sample. Error bars represent standard deviation based on three replicates.

Figure 6

Knockout of the Physcomitrella PpGT47A gene does not cause a significant change in cell wall monosaccharide composition.A) Schematic drawing of the PpGT47A gene and the knock-out construct which was introduced via homologous recombination to obtain knockout mutants in Physcomitrella. ** corresponds to a truncated version of the E4 sequence and * corresponds to a truncated version of the 3^′UTR sequence which are indicated in the upper genomic sequence. B) Monosaccharide analysis of gametophore cell walls from Physcomitrella wild-type and 4 independent gt47a knockout mutants. Data were obtained from 3 technical replicates of two pooled samples for the wild-type and knockout Ppgt47a:3, and three pooled samples for knockouts Ppgt47a:2; Ppgt47a:5 and Ppgt47a:11 .