Mutations in Glucan, Water Dikinase Affect Starch Degradation and Gametophore Development in the Moss Physcomitrella patens

Abstract

The role of starch degradation in non-vascular plants is poorly understood. To expand our knowledge of this area, we have studied this process in Physcomitrella patens. This has been achieved through examination of the step known to initiate starch degradation in angiosperms, glucan phosphorylation, catalysed by glucan, water dikinase (GWD) enzymes. Phylogenetic analysis indicates that GWD isoforms can be divided into two clades, one of which contains GWD1/GWD2 and the other GWD3 isoforms. These clades split at a very early stage within plant evolution, as distinct sequences that cluster within each were identified in all major plant lineages. Of the five genes we identified within the Physcomitrella genome that encode GWD-like enzymes, two group within the GWD1/GWD2 clade and the others within the GWD3 clade. Proteins encoded by both loci in the GWD1/GWD2 clade, named PpGWDa and PpGWDb, are localised in plastids. Mutations of either PpGWDa or PpGWDb reduce starch phosphate abundance, however, a mutation at the PpGWDa locus had a much greater influence than one at PpGWDb. Only mutations affecting PpGWDa inhibited starch degradation. Mutants lacking this enzyme also failed to develop gametophores, a phenotype that could be chemically complemented using glucose supplementation within the growth medium.

Figure 1

Analysis of putative GWD encoding sequences from Physcomitrella. (a) Predicted exon-intron structure of the five identified loci. (b) Amino acid sequence within the active sites of the predicted proteins in comparison with those from Arabidopsis. (c) Domain structure within the PpGWD proteins. PLN02784, CBM20, Pyruvate phosphate dikinase (PPDK) and PEP synthase domains are shown at the approximate sites that they are found within the polypeptides. The length of the figures represent the relative number of amino acids (aa) present in PpGWDa (1420aa), PpGWDb (1415aa), PpGWDc (989aa), PpGWDd (1170aa) and PpGWDe (1148aa). Black circles denote the C-termini.

Figure 2

Phylogeny of GWD sequences from various plant species. (a) 50% majority rule consensus tree from partitioned MrBayes analysis of the full DNA data set of Archaeplastid GWD/PWD sequences, showing phylogenetic placement of determined P. patens (bold) paralogs. The three numbers listed above branches are Bayesian posterior probabilities (PP) for the full, partitioned analysis, PP for a partitioned analysis with a region of uncertain homology excluded, and PP for the recoded RY analysis. Numbers under each branch are RAxML likelihood bootstrap support (BS) for the full, partitioned analysis and BS for a full, partitioned analysis with a region of uncertain homology excluded. Dashes indicate support for a branch <50%, or where the branch was not present in a particular analysis. Un-numbered branches received maximal support in all analyses. The scale bar is in substitutions per site. (b) 50% majority rule consensus tree from partitioned MrBayes analysis of the full AA data set of Archaeplastid GWD/PWD sequences, showing phylogenetic placement of determined P. patens (bold) paralogs. The two numbers listed above branches are Bayesian posterior probabilities (PP) for the full analysis, and PP for analysis with a region of uncertain homology excluded. Numbers under each branch are RAxML likelihood bootstrap support (BS) for the full analysis and BS for analysis with a region of uncertain homology excluded. Dashes indicate support for a branch <50%, or where the branch was not present in a particular analysis. Unnumbered branches received maximal support in all analyses. The scale bar is in substitutions per site. In (a,b) the branches subtending two land plant clades characterized by specific motifs are labelled with the corresponding amino acid sequence (VFVTC and CFATC). Sequences are identified either by Phytozome AGI codes or NCBI accession numbers.

Figure 3

Examination of GWD in P. patens. (a) Protoplasts expressing PpGWDa or PpGWDb fused to GFP. Scale bar is 5 µm. (b) PCR analysis from gDNA demonstrating the presence of mutant alleles in Ppgwda, Ppgwdb or double mutant (DM) lines. Amplicons were separated on a 1% (w/v) agarose gel. λ-PstI represent λ phage DNA digested with PstI. (c) Semi-quantitative RT-PCR analysis of PpGWD1a, PpGWD1b or Actin expression in the wild-type (WT), Ppgwd1a, Ppgwd1b and DM experimental lines. NTC designates the no template control. Original gels are shown in Supplementary Fig. 2. (d) Glucose 6-phosphate amounts in starch from the wild-type and mutant lines. Data represent means of three independent digestions of pooled starch samples ± SEM. Letters represent groups with similar means at the 5% significance level as determined using the Bonferroni-Holm post hoc test following a one-way analysis of variance. Equal variance was determined using Levene’s test.

Figure 4

Starch and soluble sugar amounts in colonies from the WT and mutant lines. Plants were grown on BCD medium for 5 weeks under a 16 h/8 h day/night regime. Tissue was sampled at five time points over a 24 hours period and starch, glucose, fructose and sucrose determined. Data represent means of at least 5 colonies. Error bars are SEM and, if not visible, are within the symbol.

Figure 5

Growth and gametophore number in the first set of mutant lines. Colonies were established on BCD media, BCD + 0.05 M glucose and BCD + 0.05 M mannitol and allowed to grow for 5 weeks. Data represent means of at least 3 colonies. Error bars are SEM and, if not visible, are within the symbol.

Figure 6

Colony morphology of wild-type and the first set of mutant lines grown on BCD medium for 1, 2, 3, 4 and 5 weeks.